5 ms·
> There is also almost no deterrent: Britain’s police solve only 5% of crimes (and 2% of vehicle thefts) Idk how this is acceptable at all. Is the UK literally
by wagwang 1y ago
> There is also almost no deterrent: Britain’s police solve only 5% of crimes (and 2% of vehicle thefts)
Idk how this is acceptable at all. Is the UK literally the state of nature?
- multjoy 1y agoCrime statistics are difficult. England & Wales (because policing is a devolved matter in the UK) have very robust crime recording rules. Consequently, the detection rates are low because you record and close crimes where there is literally no prospect of a conviction. You compare this to, say, Japan, where an investigation only starts if it’s likely that the crime will be solved, and you have an explanation for why detections seem comparatively poor. There is also the fact that, despite TVs assertion to the contrary, that solving crime is not easy and it is also true that being able to operate a fully encrypted communication system makes it harder as you rely on mistakes. As we saw with Encro, criminal groups with Signal and modern iPhones can communicate with gay abandon if they maintain decent opsec.
- WarOnPrivacy 1y ago> that solving crime is not easy True. In a healthy society, policing is hard. > and it is also true that being able to operate a fully encrypted communication system makes it harder as you rely on mistakes. Yes. You are describing actual police work; it is how things have always been. Because this was true before robust encryption, we know encryption doesn't change the equation and can be safely omitted from your assertion. > As we saw with Encro, criminal groups with Signal and modern iPhones can communicate with gay abandon if they maintain decent opsec. Governments have never had realtime access to our communications. Humans' communications have been private for as long as there has been language. Privacy is good for us and is better than all other alternatives. Robust encryption is how we maintain that natural, neutral, healthy default. Otherwise, we're talking about gifting new, unprecedented surveillance powers to officials, politicians and their powerful allies. Massive power. Over us. At which point we are less safe.
- wagwang 1y agoOK but given that only a tiny fraction of crime is solved, then why does the UK have huge prohibitions from carrying things to defend yourself, like even pepper spray?
- dan-robertson 1y agoYou aren’t allowed to attack someone for trying to steal from you so I don’t see how this is relevant to a thread about theft.
- GJim 1y ago> You aren’t allowed to attack someone for trying to steal from you Please don't spout Daily Mail rubbish. Not only is this a demonstrably false comment, it is an idiotic one.
- southernplaces7 1y agoYou're the one whose being dishonest here, at least partly. From what I've read of self-defense laws in the UK, you can use "reasonable" force to defend yourself against attack (try measuring that precisely in the heat of a moment where you don't know your attacker's aims) but yes, you can actually be prosecuted for using force against someone trying to steal from you. Even something as shitty as a google search will show that your claim is demonstratable untrue.
- multjoy 1y ago> Yes. You are describing actual police work; it is how things have always been. No, it hasn’t. Criminals haven’t always had access to literally uncrackable encryption, both messaging and voice, in a manner that is impossible to attribute to an individual handset or terminal. Regardless of whether you’re for or against it, you have to admit that this is a boon for criminals as much as it is for everyone else. Encro failed because the Dutch got all up in their servers (and the owners, who I suspect are now dead as a result, pretended it hadn’t happened) - e2e encryption bypasses that vulnerability.
- tsukikage 1y agoHow does insurance for stolen property work in Japan? Do the insurers not require evidence that the theft has been reported to the police, as they do here? Or do those reports not form part of the Japanese crime statistics, and if so, what would those look like if they did? Or is something else going on?
- ahmeneeroe-v2 1y agoI simply don't believe this. Any one crime is hard to solve but the same criminals are doing most of the crime, so if policing were effective you'd still see closure rates. >if they maintain decent opsec. don't believe this either. no one has ever maintained decent opsec.
- potato3732842 1y agoNobody you've read about has ever maintained decent opsec. There's tons of people who started as petty criminals who made it to mid ranks, got old and got out before they got killed or got caught.
- ahmeneeroe-v2 1y ago>Nobody you've read about has ever maintained decent opsec Very good point. I do think the number rounds to zero, but yes there is a data bias here.
- vkou 1y ago> Idk how this is acceptable at all. Because the only society with a high clearance rate for crime is a police state that is very good at finding someone to blame, but not necessary the guy who did it.
- prmph 1y agoWhat is the rate for major and violent crime? I imagine the 5% includes all kinds of petty crime, no?
- wagwang 1y ago2% is for grand theft auto is crazy