Curiouser and curiouser: looking under the bonnet at researcher diversity data
Posted 26/05/2021 by 9a4fa0b2-a68f-44ca-95b6-a2b900c1471a
By Dr Robin Mellors-Bourne, Director for Research & Intelligence, The Careers Research & Advisory Centre (CRAC) Ltd.
The Royal Society also published a study by JISC specifically about ethnicity in science. Some press coverage ensued highlighting the low representation of Black researchers, in particular; there were no UK-domiciled Black applicants to the University Research Fellowship (URF) scheme the last three years.
Our analysis reveals some interesting broader findings and trends about the early-career research population, or more specifically the pool of postdocs in the UK eligible for these awards:
- The size of the pool has risen by about 10% over the last five years (i.e. there are more postdocs);
- A consistent 42% are female, although this masks strong variations by subject (lower in physical sciences and engineering, higher in biosciences);
- 13% of the UK nationals are of ethnic minority background, close to the proportion in the total UK labour force, but those of Black background are strongly under-represented (1%, compared with 3% in the total labour force);
- The pool is increasingly international – the proportion of UK nationality has been falling (five years ago it was nearly half, but now 38%) while non-European nationalities particularly have risen substantially;
- Things vary a lot by subject. There are literally a handful of UK Black postdocs in physics and chemistry combined (partly why there are no applicants for a URF). On the other hand, if trends continue a few more years, the largest ethnic group in engineering (if we include all nationalities) will not be white.
That progressive internationalisation has had big effects. Many of the incoming young scientists are from Asia, mostly men. This in turn affects the total proportion of women (relatively static around 42%). Looking more closely, despite our efforts to promote progression of women in science research, both the number and proportion of UK-domiciled women have fallen very slightly over the last five years (44% of UK-domiciles, in 2018/19).
Effects on ethnicity are more complicated still. If we look at UK nationals, the proportion from ethnic minority groups has risen slightly (from 11% in 2013/14 to 13%), but numerically no increase at all because the total proportion of UK nationals pool has fallen.
But if we consider the ethnicity of all nationalities, a much higher proportion (about 30% in 2018/19, up from 24% five years earlier) can be reported as of minority ethnic background, driven by that inward mobility of (mostly male) researchers from Asia. However, the total with a Black background remains a lowly 2%, even across all nationalities. Few Black science researchers come to the UK from overseas, so the under-representation of Black researchers remains however you look at it.
Increasing international diversity also has other effects – very few internationally mobile researchers are disabled, very few work part-time, and a very high proportion are at research-intensive institutions. These intersections with nationality mean that an increasingly international workforce could be seen as having reduced diversity in some other respects.
As this workforce becomes more international, what ethnicity data should we report? Research funders increasingly report the ethnicity of applicants and awardees irrespective of nationality. The rising proportion of funding they report going to ethnic minority scientists could just reflect more international awardees, not that they are funding more UK scientists in under-represented ethnic minority groups. Does that matter?
HESA collects ethnicity data for HE staff of all nationalities, presumably because the Equality Act applies to all employees. But when it comes to students, including at doctoral level, it only records ethnicity for UK domiciles. So we need to be very careful if we compare the ethnicity of doctoral students and postdocs (to consider whether that progression is inclusive or not). We can only do it for those of UK nationality, yet that’s a decreasing minority of the total researcher workforce.
If you are into data, these ‘under the bonnet’ differences are intriguing, but are often invisible at the levels reported. I would contend we need to work out what data to collect from whom, and be precise and consistent in what we report, if we really want to understand the ethnic and racial diversity of researchers and whether we are making progression more inclusive.”
