What isn't coronavirus (for which we won't have undeniable proof it has a significantly higher mortality rate than the common flu until reliable serological tests are deployed on a large scale) will be mostly victims of the measures adopted by governments to tackle the pandemic. It's those who:
Excess mortality due to displacement from hospitals is one of the cited fears of allowing local medical resources to be overwhelmed, even if those deaths are not directly attributed to the virus.
There are excess deaths in areas without official declarations of an epidemic, such as some numbers coming out of Jakarta.
Overall death rates in many locales that still had spare hospital capacity have been noted as being higher. Some of this could be from people staying away, but some could be from the virus as well. In the absence of a viral contribution, the psychological reaction of the population that leads to the excess deaths seems to have a similar lag time to match the latent period of the virus. That seems suggestive of a more concrete linking mechanism.
However, we do have sampling events with known clusters. Ships, nursing homes, factories, meat plants, etc. With 60 or more per sample population, with some rising into the hundreds, we can start making statistical measurements of the behavior of the virus. We're not seeing a 50x asymptomatic ratio when a choir of 60 people sees at least 2 deaths and half are hospitalized. Various meat plants are seeing multiple fatalities and dozens sickened out of hundreds of employees on the line--while the necessary ratio of asymptomatic workers for some of these studies wouldn't allow for more than a few of them to even show symptoms.
Why would these known specific clusters not conform to the alleged order of magnitude greater number of asymptomatic people?
It seems like many of these may at some point resolve down to the ~1% range, which is generally consistent in regions whose health systems aren't overwhelmed.
At the upper end of these asymptomatic estimates, why would clusters be detectable at all? To that point, is the assertion that contact tracing employed by countries like South Korea is useless? That starts to happen if only 1/50 or fewer people with the virus are even noticed for the purposes of initiating a trace.
- Died of hunger or malnourishment and/or got homeless because after losing their job they couldn't afford food and/or housing;
Possible in poorer regions, but their figures are spotty and not the ones being debated with regards to the 10x-50x infection rate claims in western countries.
If people are arguing lockdowns are infeasible for locations that cannot provide sustenance or support, that can be made. Using studies whose error bars are the same size as their results and do not square with the measured behavior of known clusters is associating their position with bad data.
- Died of stopping a recurrent medication because they became afraid to leave the house to go to a pharmacy and/or couldn't afford the drugs anymore after losing their jobs;
- Died of the increased criminality generated by both the convenience of the streets now being deserted and the unemployment spike.
The 1-2 week lag being inferred for the virus and its mortality seems like a poor fit for many medications with a month or multiple months per fill.
Excess criminality is a difficult thing to measure. There are worries of under-reporting in some cities like Miami, where there's been an abrupt drop in homicides.
However, I haven't seen claims of a rise in homicides at the levels of excess mortality, and many of the opportunities for deaths due to violence have been reduced in areas that have prevented large gatherings in drinking establishments.
Social distancing has reduced opportunities for many forms of non-premeditated violence, and non-violent deaths due to workplace injury or traffic accidents have also been reduced. Vehicular deaths are one of the leading causes of mortality in general, so reducing those even slightly can have measurable effects.
For many of these scenarios in modern economies, the near and mid-term levels of disaster are heavily predicated on what policies have been chosen to deal with such disruptions. If the choice is made to not mitigate the impact of recommended actions based on decades of epidemiology study, it's not on the science or even the virus.
The US alone will be looking at over 30M unemployed this week. I hope no one's naive to the point to think this won't make crime rates soar within months, and there are always victims to higher crime rates. Especially in a country where you find a gun in every corner.
The people whose deaths are being evaluated for the death rate have already died prior to this hypothesized crime wave six months from now. The infectiousness and biological effects of the virus wouldn't discriminate on those metrics.
There is a raft of responses and staged opening of less-affected sectors to avoid desperation conditions, and whether authorities or society do not undertake them is a separate question from what the impact would have been from the expected health system overload.
At some point we'll have to wonder if the amount of people we're saving with the lock-downs is higher than the victims of the same lock-down.
Domestic incidents do appear to be significantly higher. Deaths that have been attributed to violence or unnatural causes thus far haven't been recorded to the level of the excess mortality in areas with significant infection like New York.
There hasn't been evidence that most regions with resources on the order of western countries have reached the point where more people are dying due to the stressors related to lock-downs or the milder safer-at-home or social distancing measures.
The biggest differentiator right now is the number of Covid-19 victims are being shouted out by every news outlet every day and night, while the others are so far silent in comparison.
I think a rise in murders on the order of tens of thousands of people in the space of a month would make a good news story, if such a rise happened.