4 ms·

That would be all sorts of counterproductive, starting with the simple fact that maybe the person had a reasonable 20%-likelihood guess and considered that high

That would be all sorts of counterproductive, starting with the simple fact that maybe the person had a reasonable 20%-likelihood guess and considered that high enough to bring police into it. Move just one step up and you've got people who are almost certain, but have a doubt, and since they'll get fined/punished if they're wrong then they don't call anyone, so the kid keeps being beaten and no cops ever know about it. Move another step up, and you've got more parents who would otherwise not have beaten their kids start falling down the slippery slope because no one is willing to face the potential consequences if they wrongly report a kid-beater. That's just the first three nodes of a N>10^8 causal graph. Wasting police time is an issue, yes. It would be even more of a waste of time trying to establish whether this person had reasonable cause to call the police or was just trolling. Not to mention the other consequences I touch on above.

drdeca 12y ago
Well, what if the cost if one is wrong is small enough to not be a large loss, and there is a reward if one is right to account for the risk? Suppose you want to incentivize people to report if they are at least 5% confident that there is a problem. If you choose some small cost (say, $5 for example) if one is wrong, then if you assign a reward which is at least 19 times the cost if one is wrong, then the expected monetary value of reporting (in addition to the benefit from/of being altruistic) is positive when one is at least 5% confident that there is a problem. (The ratio, where p is the minimum level of confidence you want people to report if they have, between the cost and reward, is (1-p)/p .) One might question whether creating a monetary incentive scheme like this would still cause problems (akin to the expiraments about kids being monetarily rewarded for quality of artwork in schools decreases motivation or something like that), and that does seem like something that could be the case, But one can set up a system that doesn't just disincentivize reporting when one has any doubt (as a simple fine for false reports would do), but still disince tivizes sufficiently uncertain reports. That doesn't mean it would be a good idea though. I don't know if it would be.
- DaFranker 12y ago
  Yeah, some talented thinkers could come up with some pretty good schemes. The real problem lies in the numbers: To put it simply, the cost (salaries + funding) to research this issue and evaluate a good scheme and then implement it could be much higher than several years' worth of "noisy" police calls. Add this up with the opportunity costs (what those thinkers/researchers aren't doing while they're busy with this) and it's probably all the way down to "Will Do This Only If My Next Election Depends On It".
danielweber 12y ago
When you bring in the police, you don't know what will happen. I've had someone call the cops on me because my son had a meltdown. I'm white so I just had a nice conversation with the cop, and I didn't get shot. But there are certainly people for whom even a random encounter with the cops has a good chance of turning out very badly.

drdeca 12y ago

Well, what if the cost if one is wrong is small enough to not be a large loss, and there is a reward if one is right to account for the risk? Suppose you want to incentivize people to report if they are at least 5% confident that there is a problem. If you choose some small cost (say, $5 for example) if one is wrong, then if you assign a reward which is at least 19 times the cost if one is wrong, then the expected monetary value of reporting (in addition to the benefit from/of being altruistic) is positive when one is at least 5% confident that there is a problem. (The ratio, where p is the minimum level of confidence you want people to report if they have, between the cost and reward, is (1-p)/p .) One might question whether creating a monetary incentive scheme like this would still cause problems (akin to the expiraments about kids being monetarily rewarded for quality of artwork in schools decreases motivation or something like that), and that does seem like something that could be the case, But one can set up a system that doesn't just disincentivize reporting when one has any doubt (as a simple fine for false reports would do), but still disince tivizes sufficiently uncertain reports. That doesn't mean it would be a good idea though. I don't know if it would be.

DaFranker 12y ago
Yeah, some talented thinkers could come up with some pretty good schemes. The real problem lies in the numbers: To put it simply, the cost (salaries + funding) to research this issue and evaluate a good scheme and then implement it could be much higher than several years' worth of "noisy" police calls. Add this up with the opportunity costs (what those thinkers/researchers aren't doing while they're busy with this) and it's probably all the way down to "Will Do This Only If My Next Election Depends On It".

danielweber 12y ago

When you bring in the police, you don't know what will happen. I've had someone call the cops on me because my son had a meltdown. I'm white so I just had a nice conversation with the cop, and I didn't get shot. But there are certainly people for whom even a random encounter with the cops has a good chance of turning out very badly.