Formal scientific institutions took a battering during the pandemic, and deservedly so. From the wildly inaccurate predictions of SAGE modellers to the denial of natural immunity by signatories of the John Snow Memorandum, ‘Science’ (uppercase ‘s’) has not had a good three years.
A particularly striking illustration of this is citation patterns in the scientific literature. If things were working well, the best studies would get cited the most. Unfortunately, that appears not to be the case: citations have flowed disproportionately to studies that uphold The Narrative.
In June, 2020, researchers from Imperial College London (including our old friend Neil Ferguson) published a paper in the prestigious journal Nature titled ‘Estimating the effects of non-pharmaceutical interventions on COVID-19 in Europe’.
They concluded – on the basis of a complex model-fitting exercise – that lockdowns had saved the lives of 3.1 million people across 11 European countries. That’s right, 3.1 million lives saved, and during the first three months of the pandemic alone.
Doesn’t very plausible, does it? After all, Sweden didn’t lock down, and they saw about as many deaths – or even fewer – than the countries that did lockdown. So how did the researchers get to the figure of 3.1 million lives saved?
As Philippe Lemoine notes, they just assumed that death numbers would have been far greater in the absence of lockdowns, and then took the difference between those numbers and the ones that were actually observed, and concluded the difference was due to lockdowns. Okay, but what about Sweden?
Well, the researchers fit a model in which the effect of different interventions could vary from country to country. And while Sweden didn’t have a lockdown, they did have a ban on public gatherings (of more than 500 people). So in the researchers’ model, Sweden’s ban on public gatherings ended up having the same impact as lockdown in all the other countries.
They were effectively claiming that, in France, Italy, the UK etc., lockdowns succeeded in preventing hundreds of thousands of deaths, but in Sweden, the same effect was achieved by simply banning public gatherings. Like I said, not very plausible. In fact, it’s preposterous.
The paper has been heavily criticised. However, that hasn’t stopped it being cited 2779 times! Most researchers don’t get that many citations in their entire career, let alone on a single paper.
Now let’s look at the number of citations accrued by papers finding that lockdowns didn’t have much effect.
Simon Wood’s paper ‘Did COVID-19 infections decline before UK lockdown?’ concluded that “infections were in decline before full UK lockdown”. It has been cited a total of 40 times (across two different versions).
Christian Bjørnskov’s paper ‘Did Lockdown Work? An Economist’s Cross-Country Comparison’ found “no clear association between lockdown policies and mortality development”. It has been cited a total of 57 times.
Eran Bendavid and colleagues’ paper ‘Assessing mandatory stay-at-home and business closure effects on the spread of COVID-19’ did “not find significant benefits on case growth of more restrictive NPIs”. It has been cited a total of 205 times.
Christopher Berry and colleagues’ paper ‘Evaluating the effects of shelter-in-place policies during the COVID-19 pandemic’ did “not find detectable effects of these policies on disease spread or deaths”. It has been cited a total of 68 times.
While none of these papers is perfect, they’re all vastly more rigorous than the Imperial College study published in Nature. Despite this, none of them has garnered even 1/10th as many citations as that study.
Something has gone seriously wrong when a flawed study gets almost three thousand citations, while more rigorous studies only pick up a few dozen. As to what explains this disparity, I can only speculate that most scientists haven’t come to terms with the fact that the ‘experts’ dropped the ball.
Profanity and abuse will be removed and may lead to a permanent ban.