Around the world deaths from COVID-19 are probably being significantly under-counted. The Economist makes its coverage of COVID-19 available to non-subscribers - this data-rich post shows how across many economies deaths have exceeded statistical norms. That is to be expected in a pandemic, but also that there is an excess over the number expected, plus the number categorised as COVID-19, and that those deaths follow the same pattern as COVID-19 deaths - i.e. that they peak at the same time. Therefore these are highly likely to be COVID-19 or at least strongly related to COVID-19: such as death from another cause, which may have been averted had the person sought hospital treatment, or had hospital treatment been available but for the pandemic. Much more detail on methods for effectively comparing excess death rates is available in this article from Our World in Data by Max Roser: https://ourworldindata.org/covid-excess-mortality
A weakness of this article is the lack of data from countries that have more successfully managed the pandemic, except for Norway, see below. China and Vietnam may not publish their data, or it may not be considered as reliable by The Economist. But I am sure South Korea, Japan, and Taiwan all have good data - it would be interesting to see how that compares. New Zealand data is also missing, but a good article from Charlie Mitchell and Michael Day at Stuff.co.nz has that information: click here to read more. Like Norway, in The Economist's analysis we have had fewer deaths than expected - meaning that our COVID-19 response has not just squashed the pandemic, but also other deaths: road deaths, seasonal flu, industrial accidents, and so on. Of course, downstream effects on the economy, debt, and deferred medical treatment (especially from the earlier, more restrictive lock-down) may still emerge in the coming years. This is far from over.