EWJ August 62 2025 web - Journal - Page 47
Should DEI be
on the Risk Register?
I lunched with an old friend recently. He runs a small consultancy and told me of a phone call
from the regulator regarding their BAME staff. The caller seemed to suggest that whatever his
tally was, it wasn’t enough. I thought I would test this approach.
I have modelled a small consultancy that recruits two
people every year for four years (ignore exits for now).
Let’s assume that there are four applicants each year.
Let’s also assume that BAME candidates from the
population from which I am recruiting amount to
25%, and, furthermore, I am using an agency such as
Rare Recruitment, that supplies three White candidates and one BAME one each year. I will therefore
recruit either two White ones or one White and one
BAME. I assume furthermore that we obey the law
and recruit on merit – and merit is evenly distributed
among races. This can be simulated by random
“blind” selection which is often used.
However, as far as BAME recruitment goes, this
recruitment market is serving both racial segments
well, as the total of BAME recruit comes out at just
under 25% (actually 23.25%). However, if I run the
simulation 100 times, the average comes out at 25.09%
BAMEs. Here, there are on average 36 firms that
appear to be “underperforming”.
I then ran the same exercise doubling both the
recruitment numbers (to four a year) and the period
(to eight years). For one run, we get 44 underperforming groups but an average BAME recruitment
rate of 25.23%. For 100 runs, the averages are 40.9
and 25.904%.
So after four years, I will have interviewed sixteen
people and recruited eight.
We can carry out similar work for other batch sizes,
offer rates and periods, but from the above it seems
clear that asking individual firms to meet either
quotas or targets is a bad idea. Moreover, our
“SuperBAME” firms above will have no idea why they
are “SuperBAME”!
What is my firm likely to look like, BAME – wise?
First, I note that, for each year, the probability of
recruiting two White ones is exactly the same as
recruiting one White and one BAME. I can prove this
by labelling the White ones 1,2, and 3 (and the BAME
one B). Then there are three White possibilities (1-2,13, and 2-3), and three Mixed ones (1-B, 2-B, and 3B). 50/50 chances of each.
Far better is to monitor the aggregate position and to
promote recruitment on merit.
Now, the regulatory concern is that some races are
underrepresented – which concern is often justified
by representation in staffing levels – but changes in
the supply population need to be monitored. For example, the proportion of Black graduates with better
degrees (II(i) and above) has steadily increased since
the start of this century – (although not so well for
those with firsts). How have the changes filtered
through?
You can see that, on average, firms like this will have
25% BAME staff, ie, two out of eight recruits. However, clearly, some will have fewer, and some will have
more. What will the spread look like?
I carried out a random simulation of 100 such firms,
and found the following spread, see below.
This is interesting. It shows that only just over a third
of firms fit the average. 27% are over, and 36% are
under.
Zealotry comes with a superficial review of the
situation. Activists are often keen to promote “antiracism” – which is often nothing more than “antiWhite racism” – with justifications such as “atoning for
our colonial past”. Such promotions are often presented as “contextual recruitment” – giving credit for
eg, family circumstances. This has been expressed as
“Three C’s at A level can be regarded as three B’s, due
to family circumstances”. Whether the recipient of this
credit will perform as well as the counterpart they
have replaced is moot. This raises interesting issues
for any employment lawyers addressing complaints
that recruitment was “preferred race” based rather
than merit based.
Accordingly, without doing the above work, the
regulator is likely to think “Gosh! There are 36% of
firms with too few BAMEs – and 10% without any at
all! I’d better ring them up and encourage them to
change their ways.”
The other interesting factor is the narrative often
heard “I’m the only person in the firm that looks like
me”. In our full simulation, we find there are, on
average, 26 firms out of 100 where this occurs. For
recruitment based on merit, this is impossible to avoid.
Number of BAMES in a Firm:
4
3
2
1
0
Total
Number of Firms:
5
22
37
26
10
100
EXPERT WITNESS JOURNAL
45
AUGUST/SEPT 2025
