In my youth, I was skeptical of things I could not see. Like community social health. Not just physical health, but social health, and not just of individuals, but of communities. But now that I am older and can see more, I am convinced: communities exist, and matter. Not just very visible things like jobs, parks, houses, and stores. But harder to see coalitions, cultures, and norms that influence how people feel about and treat each other.
It seems plausible that such a program which managed to *stay* secret could be effective in ways that an open one could not. The risks significant though, as "effective" cuts both ways and secrecy invites abuse.
Perhaps Robin is thinking that secretive such efforts are going to happen regardless, and he's hoping to give the well-intentioned subset of those a leg up.
Dunno... If one really wishes to improve the social health of communities, the most effective way would be to spend effort on improving the social health of *their* community. But the well-educated, mostly-well-paid, semi-academic pointy head types that tend to be readers of thoughtful blogs like this (and I definitely include myself here) tend not to put in time on local politics, local religious organizations, local charities, et al. We're citizens of the world, but not citizens of our towns.
There are platforms that do this or some variant of it to some extent or for the purposes of specific (visible) use cases like NextDoor, Yelp, Apartment Ratings, Zillow, Great Schools, or sites that provide data on crime incidence. If these tools and data sets were aggregated/ interpreted with machine learning, they could paint an interesting view of a community's social health.
Mapping the (invisible) signals of health, you run into subjectivity, and/or difficulty of measurement. For instance the signal of 'neighbors dropping by to say hello' may signal health to one person, intrusiveness to another. Littering is a very conspicuous negative behavior in a community, but how would a secret society member calibrate the volume of litter in one place versus another. Standards could be established (along the lines of a mystery shopping model), but how would data collection be financed? Crowdsource or expert-source the data, sell memberships? Maybe (ala Angie's List, Zagat). An affiliate model (sending a user of the data to relevant products and services) would inevitably lead to violations of conflicts of interests (i.e., local chamber of commerce advertises in exchange for higher scores).
There is indeed a literature, but it doesn't tell you how to see at a micro scale which local communities are healthy.
If my proposal were implemented and worked, you could predict local social outcomes from their ratings after you had controlled for the other usual visible parameters.
Tyler Cowen often remarks there is a literature on everything. And so of course there is a large literature on public trust, norms, social cooperation. With many striving to find ways to improve these, score communities and nations on social trust, etc, etc. Max Roser for example has good visulizations on this herehttps://ourworldindata.org/...
You post implies that secret ratings, done by a secret society, would somehow be superior to, or at least be a useful additional to, that body of work. Why? Why would something done in secret be better, or at least on the margin be a greater gain than the straightforward approach of adding to what is already being done? Likely you have a reason, but your post would be improved if you spelled that out more clearly. Tenatively I think your arguement is just the draw of a secret society is a big lure, so may pull in more talented folks into this field with a new approach. Is that it? Or more to it?
It seems plausible that such a program which managed to *stay* secret could be effective in ways that an open one could not. The risks significant though, as "effective" cuts both ways and secrecy invites abuse.
Perhaps Robin is thinking that secretive such efforts are going to happen regardless, and he's hoping to give the well-intentioned subset of those a leg up.
I agree this is a deserving related activity to consider.
Dunno... If one really wishes to improve the social health of communities, the most effective way would be to spend effort on improving the social health of *their* community. But the well-educated, mostly-well-paid, semi-academic pointy head types that tend to be readers of thoughtful blogs like this (and I definitely include myself here) tend not to put in time on local politics, local religious organizations, local charities, et al. We're citizens of the world, but not citizens of our towns.
The society would need to decide how long a stay was needed, and require stays of that length for valid ratings.
A subgroup could hijack the system to impose its ideals of how communities should look. Sort of like Russian Wikipedia.
What if a community tries gaming the ratings by making other places worse?
There's no way a casual traveller could accurately asses this.
Example: having moved to a new state two years ago, I am just now learning enough about my local social community to play a part in it.
There are platforms that do this or some variant of it to some extent or for the purposes of specific (visible) use cases like NextDoor, Yelp, Apartment Ratings, Zillow, Great Schools, or sites that provide data on crime incidence. If these tools and data sets were aggregated/ interpreted with machine learning, they could paint an interesting view of a community's social health.
Mapping the (invisible) signals of health, you run into subjectivity, and/or difficulty of measurement. For instance the signal of 'neighbors dropping by to say hello' may signal health to one person, intrusiveness to another. Littering is a very conspicuous negative behavior in a community, but how would a secret society member calibrate the volume of litter in one place versus another. Standards could be established (along the lines of a mystery shopping model), but how would data collection be financed? Crowdsource or expert-source the data, sell memberships? Maybe (ala Angie's List, Zagat). An affiliate model (sending a user of the data to relevant products and services) would inevitably lead to violations of conflicts of interests (i.e., local chamber of commerce advertises in exchange for higher scores).
Cool idea, just spit-balling here.
sounds insidious as all heck
In my youth, I was skeptical of things I could not see.
But you believed in God! (Or was that an earlier phase of youth?)
There is indeed a literature, but it doesn't tell you how to see at a micro scale which local communities are healthy.
If my proposal were implemented and worked, you could predict local social outcomes from their ratings after you had controlled for the other usual visible parameters.
It's not immediately clear to me why I should trust a health assessment by a secret someone.
Tyler Cowen often remarks there is a literature on everything. And so of course there is a large literature on public trust, norms, social cooperation. With many striving to find ways to improve these, score communities and nations on social trust, etc, etc. Max Roser for example has good visulizations on this herehttps://ourworldindata.org/...
You post implies that secret ratings, done by a secret society, would somehow be superior to, or at least be a useful additional to, that body of work. Why? Why would something done in secret be better, or at least on the margin be a greater gain than the straightforward approach of adding to what is already being done? Likely you have a reason, but your post would be improved if you spelled that out more clearly. Tenatively I think your arguement is just the draw of a secret society is a big lure, so may pull in more talented folks into this field with a new approach. Is that it? Or more to it?