« Snapshots of life | Gene Expression Front Page | Garrett Hardin dead »
September 20, 2003

MORE CENSUS GLEANINGS

My earlier post on data from the 2001 UK Census provoked a lot of discussion, so I thought I would dig up a few more nuggets from the Census Report.

One point of interest is the level of educational and professional achievement reached by different ethnic groups. Some of the results may be surprising...

Table S117 in the main Census report for England and Wales gives details of the highest level of qualification achieved by the main ethnic groups, broken down into age groups. For present purposes the most relevant age group is probably 25-34. Most people in this group will have completed their education and started a career.

The most useful figure may be the proportion of each ethnic group which has qualifications at or beyond GCE A-level (roughly equivalent to successful High School graduation in the US). In the jargon of education bureaucrats GCE A-level, along with some other qualifications, is described as ‘Level 3’. Beyond this are ‘higher education’ Levels 4 and 5. Levels 4 and 5 include university degrees and diplomas but also professional qualifications such as qualified teacher, nurse, or accountant. The Census gives the percentage of people who claim to have Level 3 or Level 4/5 as their highest qualification, so to identify the proportion who have reached at least Level 3 it is necessary to aggregate the figures in the Census for Levels 3 and 4/5 (still with me?)

So here are the percentages of people aged 25-34 in the main ethnic groups in England and Wales who have achieved (a) at least Level 3 (i.e. Census figures for Levels 3 to 5 combined), or (b) Level 4/5, in rank order:

Ethnic group.....................Level 3 + ............Level 4/5
Chinese............................68.4...................63.4
Indian.................................55.7..................48.9
Black African....................55.6..................45.4
Black Caribbean..............34.8..................26.1
White British.....................34.4..................25.8
Pakistani...........................33.9..................27.7
Bangladeshi......................24.6..................18.9

It will be seen that in both columns Chinese, Indian and Black African are in the top three places (in that order), and Bangladeshis at the bottom. For the other groups the rank order varies but the percentages are all quite close together.

The high position of Chinese and Indian people comes as no surprise, but the position of Black Africans (nearly as high as Indians) may be embarrassing to those who believe that Black Africans have an average IQ some two standards deviations below the White norm! Of course, African immigrants to the UK are not a random sample of the African population, and they may come largely from the upper end of the African ability range, but even so, the proportion is strikingly high.

The low ranking of ‘White British’ should not be especially surprising. Young white English working class people have never had much enthusiasm for paper qualifications, and they often prefer to leave school and start earning money (or having babies) as soon as possible. Successive Governments have tried to persuade or bully them to do otherwise, but with only limited success.

Of course, there may be suspicion that some of these figures are inaccurate or misleading. I can’t rule this out, but one way of checking it is to look at the jobs people are actually doing. Table S109 in the Report gives the occupations of people aged 16-74 who were working in the week before the Census date, broken down by ethnic group. I have calculated the percentages of each group who fall into the combined categories Managers and senior officials, Professional, and Associate Professional and Technical. The rank order is as follows:

Chinese..................46.4
Indian....................44.4
Black African.........41.1
White British..........39.3
Black Caribbean....35.7
Pakistani................32.2
Bangladeshi...........27.4

It will be seen that the differentials are narrower than in the first table, and the rank order of White British, Black Caribbean, and Pakistani has changed, but bearing in mind that the figures cover a wider age group, the correspondence is reasonably good. It would also be useful to compare the figures for 25-34 year-olds with those of the preceding age group (16-24), which are as follows:

Ethnic group..................Level 3 + ............Level 4/5
Chinese..........................55.8.......................25.0
Indian..............................47.6.......................19.5
Black African..................35.7.......................12.9
White British...................31.8.......................10.4
Pakistani.........................30.4.......................10.9
Bangladeshi...................27.2.........................9.1
Black Caribbean...........24.6.........................8.2

It will be seen that in comparison with the 25-34 age group, the relative performance of Bangladeshis is strikingly better. They have overtaken Black Caribbeans and are not far behind Pakistanis. I suspect that this is because the younger Bangladeshi cohort consists mainly of people born and educated in the UK. In contrast, the relative position of Black Africans and Black Caribbeans in this age group is noticeably lower than in the older cohort.

Although Black Africans are still just ahead of ‘White British’, the gap between Black Africans and Indians is much wider. Three possible reasons occur to me:

- the older cohort of Black Africans, mainly first-generation immigrants, are a
more highly selected ability group, whereas the younger cohort are ‘regressing to the mean’

- the younger cohort of Black Africans have assimilated some of the habits and values of the White British and Black Caribbean groups in their localities (predominantly inner cities)

- Black Africans in fact take longer to achieve their qualifications. There is a comedy stereotype of the African in Britain as an ‘eternal student’ and there is some evidence to support it. Black Africans of working age are the most likely to be currently studying for a qualification (44% compared to 17% of white people and 24% of Indian people). This is mentioned in a useful recent report, ‘Minority ethnic attainment and participation in education and training: the evidence’, by G. Bhattacharyya, L. Ison, and M. Blair. This can be downloaded as a free PDF document - go here and search for ‘Bhattacharyya’.

There remains a puzzle that both Black Africans and Black Caribbeans seem to do better both in higher qualifications and in their careers than would be expected from their school performance. Bhattacharyya et al. note that ‘on average, Black... pupils perform less well than White pupils throughout compulsory schooling’. This may be seen in the proportions of the ethnic groups who achieve good GCSE results at age 16 (5 or more passes at grades A-C):

Ethnic group....................% achieving good GCSEs in 2002
..............................................boys..........girls..........average*
Chinese..................................70..............77.............73.5
Indian......................................58...............70.............64
White British...........................46...............57.............51.5
Bangladeshi............................40...............50.............45
Pakistani.................................34..............46.............40
Black African..........................29...............44............36.5
Black Caribbean....................23..............38..............30.5

[* my source does not give figures for boys and girls combined, so I have just taken an average of the two figures. ]

This provides welcome confirmation that young Bangladeshis are catching up, and in fact have overtaken Pakistanis. But both Black Africans and Caribbeans are performing badly. For what it is worth, Black children in the UK also have somewhat lower average IQ than Whites (though the differential is smaller than in the US (see N. J. Mackintosh, IQ and Human Intelligence, OUP, 1998, p. 149). I don’t know of any figures specifically for Black Africans in the UK.

A partial resolution of the paradox is that although more Blacks go on to ‘higher’ qualifications, they do less well at them. Blacks are accepted by universities with lower entry qualifications, they are more likely to be studying below degree level, they are usually attending less prestigious institutions, they take longer to get their qualifications, and when they get them they are of a lower standard. (See Bhattacharyya et al for all this.)

This still leaves the question why people who are relatively unsuccessful at school persevere in seeking higher qualifications, and eventually get them, despite the time and effort involved. Part of the explanation may be that immigrant groups tend to be more highly motivated and ambitious - after all, they or their recent ancestors took the bold step of emigrating from their homelands, and they may transmit their motivation to their children, whether by example, instruction, or heredity. It will be interesting to see how the patterns unroll over the next ten years or so.

Posted by David B at 04:37 AM




a few points....

a majority (one person told me 90%) of the "Bangladeshis" in England are from the Syhlet district-so they are an atypical sampling regionally (my impression is that they are far more clannish and backward). additionally, richard lynn stated in a report on IQ among british ethnic groups that the somewhat lower scores of bangladeshis and pakistanis compared to whites was "due to lack of english fluency." if you know lynn, he seems not to be one to err on the side of culture/environment, so the fact that he stated this might be a pretty good indicator that it was pretty obvious. i know from talking to relatives that live in england that the syhletis run their own bengali medium (in their particular dialect) elementary schools-so it might not be surprising that some have poor english fluency (or literacy) despite being british born.

also, as per the african baseline IQ, i think we need ti discard the 60-70 ranges that are coming out of there. both british & american blacks, who are probably around 25% white (perhaps less for americans, more for brits), score around 85 or a bit above-and i can't imagine selection bias, assortive mating, selection response, etc. could have such a big effect.

Posted by: razib at September 20, 2003 01:03 PM


"Blacks are accepted by universities with lower entry qualifications, they are more likely to be studying below degree level, they are usually attending less prestigious institutions, they take longer to get their qualifications, and when they get them they are of a lower standard."

David,

Do you have figures for Oxford/Cambridge/London School of Economics/Imperial College etc.? I have a feeling that the percentage of Black students here would be much smaller than the percentages overall might suggest. Just an educated guess.

Posted by: Peter Phillips at September 20, 2003 03:29 PM


Razib,

From personal observation, Bangladeshis and Pakistanis in Britain do very badly overall. The percentage among them at or below the poverty line or on welfare is absolutely frightening (especially since Blacks and South Asian Muslims in Britain are growing several times faster than the native white population - a combination of higher birthrates and immigration). Among the middle class they are almost non-existent.

Posted by: Peter Phillips at September 20, 2003 03:33 PM


David,

Could you tell me what percent of the UK are the chinese and indian sub-populations??

Posted by: mojo at September 20, 2003 09:39 PM


If it's really the case that "... qualifications at or beyond GCE A-level (roughly equivalent to successful High School graduation in the US).", then standards have slid dramatically downhill since my days at grammar school in England.

Posted by: PhilT at September 22, 2003 09:20 AM


David:

I think this data is flawed on several counts, and it is not possible to draw any meaningful conclusions other than the already well-known one that Chinese and Hindu/Sikh immigrants are tend to be more highly motivated to succeed than others. This is readily observable in any country that is a recipient of the Chinese and Indian diasporas.

My reasons are:

1. The figures are based on what people claim as their educational attainment. This may or may not be the same as what they actually have.

2. The figures give equal weight to British qualifications and those attained overseas. Without wishing to seem unkind it is fairly common knowledge that diplomas and degrees from many non-western countries are of variable quality.

3. The data on professions is insufficiently detailed. It seems to fly in the face of common sense that almost 40% of blacks are in managerial, professional, or technical occupations. I know that most blacks reside in Greater London and that they are over-represented in the public sector, but it seems to me that they are more likely to be found in clerical, administrative or semi-skilled occupations.

Posted by: PhilT at September 22, 2003 09:52 AM


I posted some replies to comments that got lost in the site changeover. I will try and reconstruct them later.

PhilT: you are right to be suspicious about some of the data, but where it can be checked from other sources it seems roughly correct. E.g. universities and colleges have ethnic monitoring of their own students, which confirms the high levels of black (African and Caribbean) participation - though, as I pointed out in my post, they tend to be in the less prestigious end of the spectrum

As to occupational classification, I *think* this is done by the Census staff - the questionaire doesn't ask 'Are you a professional? (etc) - but gets people to say what their job is, and the Census team then classify it. There is still room for 'job inflation', but I doubt that it is a major distortion.

Posted by: David B at September 22, 2003 11:49 AM


[The next two are Dave's recovered comments from the server change - Jason M.]

Figures on ethnic minorities in particular universities are published somewhere - I think by the Higher Education Funding Council for England.
From personal observation at Imperial College (I often use the Library there), the proportion of Chinese and South Asian students is very high (or maybe they just use the Library more???)
Figures on proportions of ethnic minorities in Britain are given in my earlier post and in the Census report which I linked to.

Posted by: David B at September 22, 2003 07:31 PM


I forgot to reply to the query about IQ.
Apart from Mackintosh's book, which is a standard reference, there is an article on 'Cognitive and educational achievements in different ethnic groups', by West, Mackintosh, and Mascie-Taylor, in J. Biosocial Science (1992), 24, pp.539-54.
The gist of it all is that some older data, like the National Child Development Survey in 1969, did show a Black-White differential of about 15 IQ points, similar to the US. But the sample included a lot of very recent immigrants, and if these were stripped out the differential was about 11 points. More recent studies, like West et al, show a differential of 5 to 10 points, depending on the test used. If anyone knows a good recent study showing a differential of more than 10 points I would be interested to know.

Posted by: David B at September 22, 2003 07:32 PM