Accounting for the Variation in the Confirmed Covid-19 Caseload across England: An analysis of the role of multi-generation households, London and time
This research was featured in an Observer article showing some of the real life stories behind our numbers.
NPI director Peter Kenway said the research demonstrated that overcrowded neighbourhoods generated more Covid-19 infections than less-crowded areas, even after taking account of where the outbreak started, local deprivation and the passage of time. “Our models show that even when you allow for the obvious factors, there is still a heightened risk to overcrowded households, especially when you have older people living with younger people”.
Using data on older people who share a home with people of working age, we published a blog last week that linked this statistic with the number of confirmed cases of Covid-19 across English local authority areas. Such a link raises policy questions about the practicability of advice on social distancing for people whose housing conditions make self-isolation impossible.
This paper takes that initial finding forward in two ways. The first is to extend the statistical research by developing a model that goes beyond correlation to measure the size of the impact of various socio-economic variables on the confirmed Covid-19 caseload. Alongside the proportion of over-70s who share a home with people of working-age, these variables include overcrowding and local area deprivation.
The second is to draw out the policy questions raised by the model. Some of them relate to the policy response to the virus over the coming months. Others concern what have become norms of housing policy, especially the standards that determine what counts as adequate accommodation. These practical questions are what matter: the statistical model which highlights the underlying issue is a means to an end.
Summary of key findings
- The proportion of over-70s in a local authority area who share a household with people of working-age is confirmed to be a significant factor in accounting for the variation in the number of Covid-19 cases across England – even when levels of local deprivation, the time since the area first recorded five cases and an additional, non-specific, “London effect” are taken into account.
- In the short term, this finding poses question about the provision for people of working-age who, if they find themselves with mild symptoms, may not wish to self-isolate at home because they cannot maintain the necessary distance from other household members.
- In the medium term and beyond, the finding challenges both the norms which have come to define what constitute an adequate standard of housing, as well as local authorities’ and others’ strategic plans for the provision of homes (for example, the London Plan) which reflect those standards.