How to collect contributions from many people for something that costs very little. In particular, a fair method of paying for a commercial mailing list.
The Negative Lottery
The problem is this: you have 100 people subscribing to a small mailing list, perhaps for entirely social purposes, so there is no "mother organisation" to provide funding. About half of the people can be relied on to contribute to running costs, which are very low. Typically, running the list on a commercial service might cost Z50 per year ("Z" is a fictional unit of currency - might be a Euro, a US dollar, or 100 yen - say the cost of an ice cream). The people are typically scattered around the world.
This is not the usual funding problem: there is absolutely no difficulty in persuading members to pay. The problem is that there is (currently) no mechanism that allows individuals to transfer very small sums of money efficiently. Typically, transferring Z1 across different currency zones might cost anything from Z10 to Z20, so if 50 people contribute Z1 each, this pays the Z50 for the mailing list, but also makes an involuntary contribution of perhaps Z500 to the banking system, which I think is a terrible idea.
This is where the negative lottery comes in. Instead of spreading the cost uniformly, with the accompanying inefficiency, you spread it probabilistically. One person is chosen at random to become this year's Patron of the mailing list. Now this means one person pays Z50 plus the bank charge of say Z10. But this really isn't an unbearable amount, and the banks have been deprived of perhaps Z490 of income. What would be a bit unbearable would be if, by chance, the same person "won" every year. So you simply have a rule that after becoming Patron, a person is exempted from entering the lottery for say ten years.
That's all there is to it, really.
The FAQs of the matter
Q: Will it really work?
A: Yes A small mailing list I belong to has done this very successfully for a couple of years. The first time a number of us managed to get together in an Indian restaurant in Tokyo, and asked the waitress to pick the name out of a hat. Obviously being a bit public is a good idea, to avoid even the possibility of a suspicion of cheating, but the following year we were unable to arrange this, so the list-owner simply got one of his kids to pick from a hat.
Q: Isn't there a catch?
A: No. Well, only for the banks.
Q: Isn't this gambling? Isn't it illegal? (... immoral, etc.)
A: Interesting. The Law has a habit of viewing anything probabilistic in an erratic way. Fundamentally, gambling and insurance are the same thing; you exchange a definite loss, the stake or the premium, in exchange for a probabilistic gain. If something which you can't easily predict happens, you get some money. The difference is that insurance is betting on something you're assumed to want not to happen. (This can backfire if, as seems to be the case in Japan, it is easy to put insurance on someone else's life when you don't mind them dying.) Anyway, the law typically taxes betting and gives tax rebates for insurance. Which this counts as, I don't know. Here there's no up front loss, only a later probabilistic small gain (i.e. you're one of the volunteers who doesn't have to pay) or larger loss. I'd be interested to hear opinions on anywhere this scheme might indeed be illegal.
Q: Could it have wider applications?
A: Perhaps. While (ordinary) lotteries have been described as a "tax on innumeracy," I've long thought they could be improved on. As it stands, 10,000 people put Z100 each in, and one person gets half a million back, the others just lose their stake. Presumably the losers are made miserable by losing; anecdotal evidence suggests that in many cases lottery winners have no idea what to do with the rest of their lives, and are also made miserable. (Of course the organiser gets the other half million, and is quite happy, thank you.)
A large-scale negative lottery could make many more people happy, I think. To enter, you stake your total assets; the chosen loser's assets are then distributed among all the other participants. So many people win a bit, and are made happy. A single person is made miserable, albeit quite significantly so. Under at least one possible interpretation, this seems to be insurance, so perhaps governments could be persuaded to give tax relief on it.
(A basically unrelated factoid underlines the difficulty of applying rational thinking in this area. In the UK, betting on horses and suchlike is permitted, but subjected to a betting tax. I can't remember the rate - perhaps it's 5%. But the punter is given a choice of when to pay the tax: either 5% of the stake, or 5% of the winnings when received. Apparently, choosing to pay tax on the stake is overwhelmingly more popular. Odd, since if the punters pay tax on 100 million pounds in stakes, the government gets 5 million. But the payout on 100 million is say 70 million, so if the punters paid on winnings, they'd collectively derive the government of 1.5 million pounds. But they choose not to. Hmm.)
Q: Isn't a free list better?
A: You're missing the point. "Free" internet services alway carry a hidden burden, usually of advertising. The discussion starts after the subscribers have decided they want to pay for a commercial, but very inexpensive, service.
Q: Is this a completely original idea?
A: I doubt it. The absence of a reasonably-priced system of transferring small amounts of money is certainly a well-known problem; traditional financial organisations like banks aren't usually sufficiently flexibly-minded even to think about things personal transfers of this sort. I did suggest a name for what we really want (ah: the "Z" really stands for Zloty) - CentiZloty Electronic Cash Handling. We could call it a CZECH!