December 23rd: return to traditional workplaces differs by country

The economic geography of cities and countries has changed as a result of COVID-19. Of course, we do not yet know what the “next normal” will actually look like, but there are interesting indications in the data that we do have, on how mobility to workplaces (Y-axis) relates to the stringency of COVID-19 restrictions (X-axis).

Even in the “next normal”, it may be that we never go back to zero on the X-axis. For example, if we continue to experience episodic and seasonal spikes in COVID-19 infections, re-introducing basic measures like masks on public transport or restrictions on large gatherings may be necessary and/or cost-effective ways of mitigating the virus’s impact overall.

The markers, connected by lines, in the chart show where each country was in April 2020 (bottom right hand markers without a country label) and again in December 2021 (up until 19th December — markers further up on the chart, with country labels) on its level of COVID-19 restrictions and people’s mobility to workplaces. Pretty much every country (with the exception of Germany and Japan) had fewer restrictions in place in December, and every single country had more people travelling to their normal places of work.

What is intriguing is that the slopes of the lines connecting the dots are quite different by country. In the Nordics, it seems that despite relatively fewer restrictions remaining in December, workplace mobility remained 20–25% below baseline (which was measured in January 2020). In Latin America and Turkey, in contrast, workplace mobility was in fact higher in December 2021 than the January 2020 baseline, despite some COVID-19 restrictions remaining.

Now, you’d be very justified in protesting that all I’ve done is connect two points in time and that this proves nothing. Quite so. [In general, my charts and blogs are not meant to prove anything. They are meant to showcase some interesting data in a hopefully relatively entertaining format and provide some commentary on what the data suggests about the world around us. I try to hit a decent local optimum in terms of costs and benefits.]

So as a bonus (it is almost Christmas, after all), I’ve provided below the version of the chart that includes all months from February 2020 to December 2021. Perhaps this is the chart I should have started with all along, but it is pretty impossible to read if you are interested in specific countries. Here, the regression lines have some more explanatory power. (All of them have p-values less than 0.01).

In this view, there are several countries for which the regression lines stay below zero on the Y-axis at the point where stringency hits zero. (You can’t see it here — the chart would have been even messier.) In other words, simply extrapolating from experience to date, workplace mobility in those countries might stay permanently somewhat subdued even if there were no COVID-19 related restrictions. Even this chart is unnecessarily crude, though, because it uses monthly data. What happens if we use daily data?

In that case, many fewer countries have regression lines that would suggest lower-than-baseline mobility when (if) all restrictions are lifted. Using daily data from March 2020 to December 2021, those countries only include: Estonia, Finland, Hungary, Ireland, Japan, the Netherlands, Norway, and Sweden. Still quite a skew towards the Nordics, then.