alt Hacker NewsThe 4000 out of 700000 number doesn't seem quite right to me. That seems like maybe an average rate being applied to a cohort that doesn't match the average population. It would be more telling to compare deaths that were suspicious and missing persons numbers than all deaths.
That said, I would assume the government would monitor statistics like this as part of standard counter intelligence operations so they could see any patterns of potential issues as they pop up.
Oddly enough this exact topic came in another thread, and CDC mortality tables [1] are kind of eye opening for those who don't realize how brief life is. Take just the 25-44 year old bracket (which is probably conservative here given we're speaking of an older cohort) and the death rate is approximately 140/100k per year. For a sample of 700k people that'd be about 980 per year, around 1800 over 22 months.
Their estimate is more than twice that, but that is probably simply because of a higher age estimation. Mortality rates skyrocket upwards rapidly. At the 35-55 bracket we're already somewhere around 300/100k, which would be around 4k. And that age bracket is probably closer to reality than 25-44. Whatever the exact figure is going to be, 4k is probably a pretty decent ballpark.
And many of those deaths are going to be unusual, because pretty much all deaths at younger ages are unusual. A bit paradoxical given how 'regularly' it happens, when you look at the scale of society.
[1] - https://www.cdc.gov/nchs/data/dvs/MortFinal2007_Worktable23r...