What kind of research should inform Covid responses?

If we agree that evidence-informed policy and practice are good things, we need to think about what kind of research gets commissioned. Some kinds of research are definitely more useful than others.

We’ve been discussing the urgent needs in Covid research with Heather Marquette (after her great April FP2P posts on this) and others –here are a few of the topics that have cropped up.

We need actual data about the Covid-changed world

We have all largely depended on hypotheses and narratives so far. But we now need new data and synthesis to aid decision making – including to help inform trade-offs and policy choices between controlling the virus and controlling everything else that can kill, including indirect effects. This includes data on comorbidity – deaths from everything else while health service use drops or as a result of the impacts of lockdown measures, not just the virus. 

This data should not be limited to humanitarian settings. CGD have a really useful initial effort at this – trying to calculate how many non-Covid deaths may result from different kinds of Covid responses, to put numbers on the trade-offs, which are otherwise being discussed in rhetorical terms. We’ve all heard them: ‘we will all die of hunger because of this lockdown’; ‘the cure is worse than the disease’.

We are fast reaching the end of the road for Covid expert opinion based on what was happening in last data rounds before Covid, and need to shift gears to investing in new data.

Some of the big quantitative research outfits are doing this (IPA, CGD, IGC, Y-RISE) – as have some qualitative researchers (BRAC, Hrishipara Diaries, Rift Valley Institute, GAGE).

To shift to this kind of remote survey work (whether qualitative or quantitative), initial experience suggests the importance of pre-existing research networks, teams, ethical frameworks, skills; trained field researchers who can readily and safely shift to phone/remote methods; and existing baselines to sample/include/compare with. 

It is very hard to kick start decent urgent research without this foundation already in place – as we have seen with Covid ‘events lists’ (eg of protests) that lack any sense of what happened previously.

New and better narratives

There are extraordinarily conflicting Covid narratives about what is actually happening in many parts of the world, but perhaps most of all in Africa – is it doing better than the North?; or facing a crisis only slower than elsewhere (WHO)?; or facing a catastrophe now?. 

It is hard to find evidence to test these. There are also a variety of CSO narratives – from the ‘seizing opportunity to deliver mutual aid’ – to impending doom for them and their work.  Again, hard to get beyond anecdotes or individual cases.

And as the amount of Covid commentary continues to soar exponentially (no sign of curves flattening there), this is only going to become more pronounced. There is an urgent need for credible, accurate syntheses of what is emerging from all this (as no-one will have the time to read it all). But the incentives in academia are all around churning out new papers, not comparing existing ones or making sense of the mountains of blogs, articles, briefing papers, and more.

The Bigfooting power of a big emergency

Process matters. In ‘peace time’ we (hopefully) had a general agreement about who should be involved in a given piece of research to ensure a proper and inclusive understanding of views and impacts. 

Whether related to gender inclusion and women’s representation, north/south, or complementary research disciplines, we imagined that retaining this diversity and representation would also see us well through a crisis – but of course this is not the case as @Peter_Evans_guv explains.

‘Bigfooting’ in journalism is when someone has been freelancing in some largely ignored part of the world, and then something big happens (earthquake, war etc). In flies the big name correspondent from Washington or London and ‘bigfoots’ the hapless freelancer.

Something similar happens with big disasters – years of careful progress in building research methods that understand the local context, listen to citizens, work respectfully with local partners etc are shoved aside by the urgency to get the money out the door. The loudest voices, or those with the most confident offer, move quickest or speak loudest (including ‘Covid washing’ opportunists who have rebadged their pet project for the crisis).

Modelling v Empiricism

Watching how and when research influences Covid policy, we have seen the power of ‘mathiness’ – things with numbers win the argument (especially if repeated often), even if those numbers are largely arbitrary/fictional.

But drilling down, there’s a distinction between two types of science going on here, as a recent piece in the New Statesman explained. One is theoretical, meaning it is based on models that attempt to predict how things will behave. The other is empirical, meaning it is based on observation and experience of how things have actually behaved.

I much prefer the empirical side – but that takes time and meanwhile, modelling exerts a hypnotic fascination on decision-makers because they think (wrongly) that it predicts the future. 

On arrival at Oxfam, when trade was the big global issue, I spent ages pointing out the bogus nature of trade modelling, which merely added mathiness to the assumptions of the liberalisers. I even got one of the father’s of Computable General Equilibrium (CGE) modelling to write a paper denouncing how it was being misused. Don’t think it had much impact, sadly.

So where do we go from here?

We can’t fill all the gaps, so where do we start? Seems to us we need several pieces to come together around:

• More observation of ‘emergent agency’ – how are poor individuals and communities responding to the disease, and the vagaries of official responses? What new forms of politics are emerging? LSE and Oxfam are collaborating on this – will post more on this in a few days.

• More qualitative research on impact and response – intersectional data on how vulnerable groups are being affected (eg increasing gender inequalities or not) and the nature and effectiveness of state/private sector/civil society responses

• More attention to fragile and conflict-affected settings where standard responses focussing on the state, private sector and CSOs are even less likely to work (eg if you’re in a bit of Myanmar that’s governed by an ethnic armed organisation!)

• More syntheses of the avalanche of research as it emerges.

Given that no institution can do all this at a global level, maybe a place to start is to identify a few countries where research networks are in place, and begin there. It will all take money of course, so a lot will depend on the ability of the big funders and foundations to move swiftly.

Any other ideas?

This post has been re-blogged from Poverty to Power.