"Every time lockdowns were introduced, Covid cases were already in decline. They'd have gone on falling even without lockdown - and then we'd have had far less of the damage lockdown causes."

1) Each of the three times the UK has introduced national lockdown measures, cases have gone from a rising trend, to a falling trend. Similar measures have had similar effects around the world. It's possible to cherry-pick some regions of the UK that did not show this pattern, but the overall trend for each lockdown is clear, as we'll show below.

2. Cases were not falling before the January 2021 lockdown. The weekly cases of coronavirus were rising across the country prior to the lockdown. See the figure below from the FT:

Indeed, the ONS infection survey shows that the infection rate in the UK on 18 December 2020 was 1 in 85, and this rose to 1 in 50 by January 2 2021.

Of course, the picture is complicated by the fact that different regions saw different levels of restrictions prior to lockdown. London for example, was moved to Tier 4 ("stay at home") on 20 December 2020. The 7-day average case rate peaked and started to fall a bit over two weeks later. Areas which were not put into tier 4 at this point generally saw cases fall a little later.

3. Cases were not falling before the national measures introduced in November 2020 either. The REACT study showed cases were rising prior to the imposition of the November 2020 measures. Following lockdown, cases were reduced by 30%:

4. Cases fell after the first national lockdown in the UK. The shaded area below marks the period of the first lockdown:

5. Cases fall when lockdowns are imposed. Whereas there are some voluntary measures taken immediately before lockdowns are imposed, analysis from Imperial College London shows that “only national lockdown brought the reproduction number below 1 consistently in the first wave”.

6**) Instead of positive tests, some point to alternative data sources, but they are not representative.** For example, some have argued that data from symptom checking apps suggests cases were turning around before lockdown measures were introduced. But these apps are reporting the self-reported symptoms of a self-selected group, rather than positive tests from the whole population, which means that they may be biased. For example, they might overrepresent more conscientious people, or people following Covid news more closely - people who are more likely to change their behaviour before a lockdown is imposed.

7. A published paper seeks to argue that restrictions have not worked, but contains a glaring flaw. A paper that appeared in January 2021, co-authored by John Ioannidis, looked at the correlation between restrictive measures introduced by government and the number of cases. Ioannidis and his colleagues found that some lockdown restrictions were even correlated with higher growth in coronavirus cases. This should have been a warning of an obvious flaw - which is that case growth and restrictions are endogenous - in other words, governments have brought in tighter restrictions when cases are higher. The argument of the paper is like arguing that "people in hospital are more likely to have heart disease; therefore hospitals cause heart disease."