Looking at lockdowns

The reopening of pubs in the UK this weekend marks an important milestone in the easing of lockdown conditions (though the Scots will have to wait until 15 July). It is 103 days since English residents have been able to take advantage of their inalienable right to drink alcohol in their local pub. This is historically unprecedented: As far back as anyone knows, pubs have always remained open. The last time there was an assault on people’s right to buy alcohol on licensed premises was during World War I when the government reckoned that drunkenness was undermining the war effort (a highly questionable assertion). This led to the imposition of limits on opening times, which lasted until the 1980s, but the pubs nonetheless remained open.

There is little doubt that the lockdown has been the most stringent imposition on individual freedoms in living memory and it has been repeated across the world, with news bulletins showing pictures of empty shopping malls and motorways that would normally otherwise be packed. In an attempt to compare the extent of the lockdowns across different countries, I am indebted to the work of academics at the Blavatnik School of Government at Oxford University who have constructed lockdown indices across more than 160 countries. The index is composed of a series of individual policy response indicators based on a range of indicators (interested readers are referred to the working paper for more detail, which can be found here).

The data are collated over the period since 1 January 2020 and chart 1 shows the maximum value of the index over the course of this year-to-date. Based on the random sample of 16 countries used here, the UK ranks fairly low down in terms of lockdown stringency. It is notable, however, that it is slightly ahead of Germany yet Germany had a very low death rate from Covid-19 infections[1] (10.8 per 100,000 of population versus 65.6 in the UK). Italy introduced the most stringent set of measures and recorded a high number of deaths (57.6 per 100,000). As has been well documented elsewhere, Sweden’s lockdown was relatively relaxed and although the death rate is still relatively high it is still below that of the UK and Italy at 52.1 per 100,000. Clearly, the maximum extent of the lockdown is not very meaningful as an indicator of the severity of the Covid-19 outbreak.

 

A more useful indicator might be the average measure of the index which also takes duration into account. But here, too, there is no ordinal ranking from lockdown severity to Covid-19 mortality rates. Chart 2 indicates that Italy’s average lockdown is significantly higher than that of Germany but mortality rates are too. To make the point more clearly, the scatter plots in chart 3 clearly indicate only a limited negative relationship between the extent of the lockdown and Covid mortality or infection rates.

This obviously raises the question of how useful have the lockdowns been as a defence mechanism against Covid-19. The trick is not to think in terms of levels but rates of change. Or to put it another way, how much worse would infection rates have been in the absence of lockdowns? Here we seem to be on safer ground for there is a very strong relationship between the extent and duration of the lockdown and the rate of infections (chart 4). In the three European examples chosen here, it is notable that Italy introduced the lockdown long before infections peaked whereas the UK clamped down quite some time after infections started to pick up, thus supporting the view that the UK was late in acting which contributed to the surge in mortality.

Lockdowns do, of course, have a significant economic cost. The lockdown indices on their own are not particularly useful and have to be augmented with other indicators such as mobility trends and specific country characteristics to render useful information. The Bank of England has done some sophisticated modelling work using the lockdown index as one input and concluded that the data across a range of countries is consistent with a decline in world GDP of around 15% over the first half of 2020 (chart 5).

Prior to this weekend the UK lockdown index was still fairly high in comparison to a number of other countries. That is appropriate given that Covid-related deaths are still high in a European comparison. However, they are well down from their previous highs and given the improvements in recent weeks, it is probably appropriate to begin a limited form of easing. But striking the right balance is virtually impossible. Some think that the infection rates are too high to justify any form of lockdown whilst others believe it should have been lifted long ago. We will only know whether the current policy stance is the right one when we can assess the trends in Covid infections.

---

[1] I use mortality rates rather than the number of reported infections because this is dependent on the breadth of the testing programme which differs across countries.