Super-spreading, mobility and crowding

I still see quite a few journalists in India refer to “super-spreaders” vis-à-vis the novel coronavirus – implying that some individuals might be to blame for ‘seeding’ lots of new infections in the community – instead of accommodating the fact that simply breathing out a lot of viruses doesn’t suffice to infect tens or hundreds of others: you also need the social conditions that will enable all these viral particles to easily find human hosts.

In fact, going a step ahead, a super-spreading event can happen if there are no super-spreading individuals but there are enabling environmental conditions that do nothing to slow the virus’s transmission across different communities. These conditions include lack of basic amenities (or access to them) such as clean water, nutritious meals and physical space.

A new study published by a group of researchers from the US adds to this view. According to their paper’s abstract, “Our model predicts higher infection rates among disadvantaged racial and socioeconomic groups solely from differences in mobility: we find that disadvantaged groups have not been able to reduce mobility as sharply, and that the POIs [points of interest] they visit are more crowded and therefore higher-risk.”

And what they suggest by way of amelioration – to reduce the maximum occupancy at each POI, like a restaurant – applies to a mobility-centric strategy the same way reducing inequality applies to a strategy centred on social justice. In effect, disadvantaged groups of people – which currently include people forced to live in slums, share toilets, ration water, etc. in India’s cities – should have access to the same quality of life that everyone else does at that point of time, including in the limited case of housing.

This study is also interesting because the authors’ model was composed with mobility data from 98 million cellphones – providing an empirical foundation that obviates the need for assumptions about how people move and where. In the early days of India’s COVID-19 epidemic, faulty assumptions on just this count gave rise to predictions about how the situation would evolve in different areas that in hindsight were found to be outlandish – and in some cases in ways that could have been anticipated.

Some modellers denoted people as dots on a screen and assumed that each dot would be able to move a certain distance before it ‘met’ another dot, as well as that all the dots would have a certain total area in which to move around. But as two mathematicians wrote for Politically Math in April this year, our cities look nothing like this:

According to this report, “India’s top 1% bag 73% of the country’s wealth”. Let us say, the physical space in our simulation represents not the ‘physical space’ in real terms, but the ‘space of opportunities’ that exist. In this specific situation of a country under complete lockdown because of the pandemic, this might mean who gets to order ‘contactless’ food online while being ‘quarantined’ at home, and who doesn’t. In our segregated simulation space therefore, the top chamber must occupy 73% of the total space, and the bottom chamber 27%. Also, 1% of the total number of dots occupy the airy top chamber, while the remaining 99% of the dots occupy the bottom chamber.

As a result, and notwithstanding any caveats about the data-taking exercises, researchers reported that Dharavi in Mumbai had a seroprevalence of more than 50% by late July while three wards in non-slum areas had a seroprevalence of only 16%.

The flawed models still can’t claim they could have been right if Mumbai’s slum and non-slum areas were treated as distinct entities. As T. Jacob John wrote for The Wire Science in October, one of the reasons (non-vaccine) herd immunity as a concept breaks when applied to humans is that humans are social animals, and their populations regularly mix such that ‘closed societies’ are rendered practically impossible.

So instead of mucking about with nationwide lockdowns and other restrictions that apply to entire populations at once, the state could simply do two things. First, in the short-term, prevent crowding in places where it’s likely to happen – including public toilets that residents of slums are forced to share, ration shops where beneficiaries of the PDS system are required to queue up, workplaces where workers are crammed too many to a room, etc.

Obviously, I don’t suggest that the government should have been aware of all these features of the epidemic’s progression in different areas from the beginning. But from the moment these issues became clear, and from the moment a government became able to reorient its COVID-19 response strategy but didn’t, it has effectively been in the dock.

This brings us to the second and longer term thing we should do: with the novel coronavirus’s transmission characteristics as a guide, we must refashion policies and strategies to reduce inequality and improve access to those resources required to suppress ‘super-spreading’ conditions at the same time.

The simultaneity is important. For example, simply increasing the average house size from 4 sq. m, say, to 8 sq. m won’t cut it. Instead, buildings have to be designed to allow ample ventilation (with fresh air) and access to sunlight (depending on its natural availability). As researchers from IDFC Institute, a think-tank in Mumbai, noted in another article:

Dharavi’s buildings and paths are irregularly laid out, with few straight routes. Based on calculations with OpenStreetMap routes and Google Earth imagery, it appears 68% of pathways and roads are less than 2 m wide. Such a dimension offers little space for air circulation, and reduces airflow relative to other, properly planned areas, and admits fewer air currents that could help break up the concentration of viral particles.

Mitigating such conditions could also impinge on India’s climate commitments. For example, with reference to our present time in history as the hottest on record, and many countries including India experiencing periods in which the ambient temperature in some regions exceeds thresholds deemed safe for human metabolism, science writer Leigh Phillips wrote for Jacobin that air-conditions must be a human right:

What would it mean to have a right to air-conditioning? Precisely, the right should be to have free or cheap, reliable access to the thermal conditions optimal for human metabolism (air temperatures of between 18 degrees C and 24 degrees C, according to the WHO). Neither too hot nor too cold. The right to Goldilocks’s porridge, if you will. New buildings must come with A/C as part of any “Green New Deal”. The aim of any programme of publicly subsidised mass retrofitting of old buildings shouldn’t be just to fuel-switch away from gas heating and improve insulation, but also to install quiet, efficient air-conditioning systems. At the scale of the electricity grid, this demand must also include the requirement that A/C run on cheap, clean electricity.

So really, none of what’s going on is simple – and when governments respond by offering solutions that assume the problem is simple are avoiding dealing with the real causes. For example, ‘super-spreading’ is neither a choice nor an event – it’s a condition – so solutions that address it as a choice or event are bound to fail. Seen the other way, a community with a high prevalence of a viral infection may be much less responsible for its predicament than the simple interaction of their social conditions with a highly contagious virus.

But this doesn’t mean no solution except a grand, city-scale one can be feasible either – only that all solutions must converge, by being targeted to that effect, on eliminating inequalities.