Imagine signing up for a social-media account that requires a small amount of money -- say $1 to $10 -- be put down as a type of security deposit. If, during your first X days of activity on the site, you flagrantly violate the terms of use (e.g. by spamming lots of other users) you lose the money; otherwise it's returned to you.
If such a system were to become commonplace, it would introduce only a minor amount of friction for most legitimate users, but it would become a great impediment to spammers and other malicious users.
However I think this is a great idea to explore for "exclusive" / "high value" forums, say for trade / special-interest groups. I'd be happy to put down a $10 stake to participate in something like Hacker News, for example, if that would help to combat astroturfing and bad behavior. (Honestly for how much I post on HN I'd be prepared to stake more.)
There's been a lot of research done in this area by the Ethereum devs under their migration from "Proof of Work" to "Proof of Stake", including research on the game-theory of "slashing conditions" (i.e. when you lose your stake). You get some cool community-policing possibilities by allowing any member of the community to collect a bounty paid for by claiming part of the forfeited stake, if they can present an example of a member breaching the rules. For something requiring review this would perhaps have less benefits, but could still be useful; rather than moderators having to watch every message go by, they would just look at the "flagged" queue, and users would now have a financial incentive to add something onto the flag queue, perhaps also having to pay a small amount to take that action as well, to disincentivise spamming "flag" entries. I'm sure you could build this all on Ethereum if you were so inclined, with "admin moderation" being a call to an external trusted Oracle service where the moderators live.
Of course there are also substantial risks in turning a system based on community goodwill into a financial game; people behave differently when they are purely optimizing for a financial game vs. when they are not.
1. They have (or are willing to use) an online payment source (credit card, paypal, etc.).
2. They have (or are willing to spend) a disposable income.
3. They trust you to get their money back.
Also, you're going to deal with a huge amount of customer service as every single person argues with you that they didn't violate the terms and deserve their money back - that's going to cost way more than any fees/interest you'd earning from the escrow.
1. What purpose would escrow serve in this example? The user trusts the social media site, and the amount is small (like $1). So there's no need for a third party to hold/control the funds and determine when they are released to one party or another.
2. A spammer can just wait until day X+1 to start spamming.
Interestingly, both are platforms for individuals to sell goods to other individuals. And both internalized the feature (no apparent 3rd party involved).
"If such a system were to become commonplace, it would introduce only a minor amount of friction for most legitimate users..."
first, that doesn't sound like minor friction, but more importantly, even seemingly small friction can lead to large abandonment, especially if frictionless alternatives are readily available.
banks love (large) escrow products by the way. super-low risk, low effort, and very reliable fee income for them.
Also, what is stopping me from spamming the site after I get my money back? like I can queue up accounts, wait for them to age up and just start spamming.
https://community.algorand.org/blog/utilizing-algorands-asc-...
https://developer.algorand.org/docs/build-apps/apps/#escrow-...
Super easy for developers to get started with too: https://developer.algorand.org/
unfortunately banks charges relating to payments govern the cost of an escrow service hence it makes more sense for larger value transactions
well, TBH both comes from the same word, not directly related. still, fun to see how the same words can end up having exact opposite meanings.