I don't think that's a good metric either. We're in an unprecedented economic and social situation, so why should we expect death rates from previous years to serve as a reasonable baseline today?
True. Examples: Less driving means less vehicular death. We probably have slightly more death due to domestic violence. More or less death from drug overdose. Less contact means less death from the normal flu.
I agree, it's not exactly a metric, but a heuristic to put official numbers in perspective. Most of the variations you say are in the direction of lowering normal numbers (not domestic violence, but crimes will be registeres as such) so we have a minimum. If official numbers are much less, it's a clear sign that they're altered.
There are several ways deaths are being covered: people dying at home or retirement homes, people with previous diseases, putting immediate cause (like stroke or heart attack) in the report.
To be fair, nationwide deaths might be the only correct metric, since it's the only thing that countries care to measure properly and don't attempt to manipulate.
narag said it is a better metric. It isn’t a perfect metric by any means, but it is one of the only fair metrics you or I could have access to when trying to compare between countries. Most other metrics are deeply flawed for that purpose.
