Increasing personal security of at-risk high-impact actorspost by meerpirat · 2020-05-28T14:03:29.725Z · EA · GW · No comments
This is a question post.
Did anybody consider providing personal security to key people with hightened safety risk as a potentially impactful intervention?
Now and then I hear about death threats against mildly prominent people that I consider doing very important work. For example this week, the most well known German virologist received a flask allegedly containing SARS-CoV2 in his mail , and earlier he reported online death threats. I wonder if there are studies on baserates of such death threats being realized, and if there is value in supporting people hire (or pushing goverments to provide) personal security earlier than they would do today.
Googling "how often are death threats realized" or "death threats statistics" didn't yield much for me, except a study that found that within "10 years, 44% of [613 tracked death] threateners were convicted of further violent offending, including 19 (3%) homicides." Of those 19 homicides, 5 were the initial death threat victim, i.e. ~1% of the death threats were later realized . I didn't see more about victim statistics, but I expect that the majority of them to be somewhat close to the perpetrators, and not public people. The death threats came from people that were convicted of their death treat, which I expect makes it even less representative for people threatening others anonymously via mail or the internet.
I suppose providing more security could be very impractical and extremely costly. Though maybe there are some low-hanging fruits, like providing safety training, limiting risk exposure (e.g. delivering groceries, removing adress details online), or hiring someone that can assess the seriousness of a death threat better than the police.
I suppose this could also be a classic case of risks that only seem noteworthy because they are so graphic and easy to imagine. Anyway, I'd be interested in what you people think.
Comments sorted by top scores.