My First GT Idea
This is what I thought about a couple of weeks after Moshe first explained to me the basic concepts of game theory. It won't save the world… but maybe it might help a bit.
Suppose you have an important problem that everyone wants to solve, and which can be solved by throwing money at it. I mean, a problem on which people already made a whole lot of research and are pretty confident that doing X,Y,Z will solve it - now all is needed is funding for X,Y,Z. I'm not sure such problems exist; for example, "world hunger" is not in this category - to much of my (limited) knowledge, this problem cannot be solved by money alone. But suppose, for the sake of discussion, that we do have such a problem, and solving it will require $M. The question is, how to gain the money?
It is long known that simply asking for donations probably won't help. I won't cite the relevant papers now, but the reasoning is as follows: when each person decides whether to give or not, he doesn't know how much will everyone else give. If the total amount that others will give exceeds $M, then his donation is useless to change the outcome, and he'll lose money by donating. If this total amount is less than $M minus whatever he's considering to give, then it's the same thing: his donation will not change the outcome, since there still won't be enough money to solve the problem. The only case where his donation will cause the problem to become solved from unsolved is if the total sum given by others will fall in the range of $[M minus whatever he considers to give, M], and the probability of that is extremely small. Therefore, the strategy that maximizes the expected revenue is: not to donate.
Now, one possible flaw in the above argument is that we assume that anything less than $M is just as good as nothing - which is, in most cases, not true at all. One of my favorite Greta Christina essays, The Harm Reduction Model Of Life, talks exactly about this: even if you cannot completely solve a problem, you can still make it better. But then, the most important question that arises is this: whether my marginal contribution is worth my effort. In many cases, it is. But in too many cases, it is not. In too many cases, if there is not enough money to solve the problem completely, the money is being used for temporary relief - which, for many contributors, is just not good enough as a stimulus. For example, if I knew that we could eliminate world hunger if I only gave all the money I had - I'd do it. Definitely. But I won't give $5 to feed the starving kids of Africa, because feeding them not only doesn't solve the problem - it doesn't even attempt to. Solving the problem would mean enabling the people of Africa to feed their own kids. That's a solution. Me, buying them another kilogram of potatoes? Not a solution; and not even worth my effort.
So, if we have the annoying case where a half-assed solution is not worth the marginal efforts, what do we do? We need $M - not less.
I propose the following: let's ask for donations with a money back guarantee. Let's have a huge campaign, saying: "This year, humanity is going to solve problem ***! Hurry up to participate! Full money back guarantee!" The campaign should emphasize the following points:
- The goal is to collect $M by a fixed date. The collection stops the moment $M is reached, and no more donations are accepted (this is important!)
- A running counter with the total amount gathered so far will be available on the Internet.
- Each person or corporation who contribute money gets a beautiful printed ticket with inscription that goes something like "I participated in the solution of *** and donated $*!" These tickets should be well protected against replication, because:
- If less than $M are collected in time, each such ticket could be returned within one year in exchange for the money contributed. This process should be as easy as possible.
Now the incentives change. Firstly, now you won't lose any money if the thing won't work - you could always get it back. Secondly, I claim that even if a person believes that $M will be reached without him, he will still choose to participate. This is where the importance of gathering exactly $M (and not more) comes into play, psychologically. The question for each person is: suppose humanity indeed solves this big problem. It will be a huge step in history. Do you want it to happen with your help or without it? Note that the contribution of everyone who buys a ticket is necessary. If you decide too late - you won't get it.
I believe that if the issue was promoted this way, then for most people it would be a dominant strategy to contribute (meaning, they would prefer to contribute no matter what other people did).
Of course, there are many problems with this idea.
First of all, the campaign itself is going to be expensive (printing the tickets, marketing, etc.), so the total amount gathered has to cover these costs as well.
More importantly, I don't know of any problems that could be solved this way - either a problem cannot be solved by money alone, or it is not important enough to make the incentives right, or it would simply require too much money to be raised.
One thing that could help, though: once you have the idea and the plan, to do an Internet poll. Simply ask people: if we do this campaign, would you give money? That way we might know if this is going to work or not.
So, questions for you:
Do you think this could work? If no, why?
If yes, for which problems attempted to be solved this way would you personally donate money?
Which problems you think could be solved this way?