Q&A: Ethnicity & Covid-19 – What the data is telling us
Thu 11 Jun 2020 | Richard Webber
There has been significant coverage on the apparent link between ethnicity and the risk of contracting coronavirus, but what does the data actually reveal?
In 2014, Richard Webber and Trevor Phillips OBE started Webber Phillips, a data analytics business specialising in a software package called “Origins” (developed by OriginsInfo, of which Webber is MD). Origins applies reference tables linking people’s names and cultural backgrounds to the examination of cultural diversity among private and public sector organisations. It wasn’t long into the pandemic before Webber and Phillips realised their software could be applied to Covid-19, amid early indicators that the UK BAME community was disproportionately at risk.
We spoke to Webber to understand the power of Origins and how the software package might help the UK Government better understand the connection between Covid-19 risk, ethnicity and occupation. In terms of what the data is telling us, Webber warns against oversimplification and says the devil is in the detail.
What is OriginsInfo and why was it founded?
The idea that a person’s name might indicate where that person’s forebears came from originated one Christmas at Experian when it was decided to create maps of the geography of each of the country’s most commonly occurring surnames. These proved very popular as Christmas presents to friends and family and a significant amount was raised from their sale.
Originally non-British names were ignored, but after I retired from Experian I began to realise that the geography of non-British names was equally revealing and, in most cases, useful – rather than just of personal interest.
How do organisations typically use Origins?
Most users of Origins are public sector organisations who want to identify how different minorities use their services and where these minorities live. For example, Yorkshire Water uses Origins to identify minorities which could be persuaded to use special containers for storing cooking oils that might otherwise be washed down the sink; a number of police forces use it to identify groups that are underrepresented among sign-ups for Neighbourhood Alert schemes; arts organisations use it to identify which minorities go to which performances and which go to hardly any; Greenwich Council used it to identify where in the borough its Nepalese community live; political parties use Origins to improve the match between the ethnicity of prospective councillors and the wards they are chosen to stand in.
Can Origins help the UK Government’s Covid-19 efforts?
Government is, of course, anxious to better understand whether minority populations are at risk and if so which ones are most likely to suffer from the virus. Unfortunately, health records categorise patients’ ethnicity only by four or five broad groups. This does make it possible to identify the differential risk of key groups such as Filipinos, Turks, Somalis, Hispanics (which it would be very easy to do if Origins technologies were to be used).
Clearly, the occupations which are at more at risk of catching the virus are those that come or have come into contact with large numbers of different people at relatively close quarters, not just in the health service but as supermarket cashiers (Tamils), taxi drivers (Muslims), restaurant staff (Bangladeshis), bus drivers (West Indians) and care workers (Nigerians and Ghanaians). Origins is a quick way of verifying the concentration of different minorities in these occupations.
What are the challenges of conducting data analysis into Covid-19 risk?
Data analysis on who is affected by Covid-19 is easy to deliver, quick, inexpensive and unobtrusive. The principal challenge is the obstacles imposed by data protection legislation on the use of people’s names for analytical purpose. Normally appropriate, in this instance, there is uncertainty over how far the protection of personal data should have priority over the protection of people’s lives, particularly among the white British community.
What is the data telling us so far?
Work we have done prior to the appearance of Covid-19 has consistently surprised us by the strong link it shows between ethnicity and occupation. Unfortunately, this work has been fragmented and we are unable to support the government with comprehensive information on this link. All we can say is that the link is strong. At present, it is not clear whether or how much of the higher rates of death among non-white British citizens is due to the minorities being employed in high-risk occupations, how much is due to genetics and diet and how much due to their experience of worse housing, poorer health care and general afflictions resulting from racism.
In our view, there is little doubt that not all non-white British groups experience disadvantage to the same degree. The days when disadvantages were based solely on the colour of your skin are long gone. Many groups that would declare themselves as white (Turks, Albanians, Romanians, Lithuanians) currently live in much worse conditions than some of the groups that would describe themselves as black or brown (Tamils, Ugandan Asians).
There has been a lot of coverage on this topic in the media in recent weeks. Are there any misconceptions that you’d like to clear up?
Dividing the world crudely into Black and White tends to jar with people’s own daily experiences. There is no doubt that on average the BAME community has a right to complain about obstacles to a straightforward life. But life is far more onerous for some communities and within every community, there are some who live far more comfortable lives than others. Public interventions need to take into account worst-hit groups but also members of those groups who live in the most disadvantaged neighbourhoods.