[fullmetal]

It is likely that the non-COVID US mortality figures depend on behaviors that the lockdown enables -- high rate of motor transportation, frequent social consumption of alcohol, increased stress from work/commuting, etc.  Hard to tell without a more detailed breakdown.  Anyway, lockdown-induced decreases in US deaths might have compensated for COVID deaths initially, up until April.

The other issue with that graph is that a very large number of people die in America compared to every other nation represented.  Those y-axis scales are unique to each nation, and the US is just off the charts by comparison.

[fullmetal]

It is likely that the non-COVID US mortality figures depend on behaviors that the lockdown enables -- high rate of motor transportation, frequent social consumption of alcohol, increased stress from work/commuting, etc.  Hard to tell without a more detailed breakdown.  Anyway, lockdown-induced decreases in US deaths might have compensated for COVID deaths initially, up until April.

The other issue with that graph is that a very large number of people die in America compared to every other nation represented.  Those y-axis scales are unique to each nation, and the US is just off the charts by comparison.