# Funding public goods

Public goods in economics are those things that are desirable, but neither rivalrous nor excludable. That is, one person’s enjoyment of the good doesn’t detract from anyone else’s enjoyment of it, and it’s impossible to prevent people who haven’t paid for it from enjoying it.

For the purposes of this article, the main example of public goods I’ll be referring to is that of free cultural works, such as open source software and public domain audiobooks.

Obviously, a lot of these public goods are being produced, sometimes for altruistic reasons, sometimes for fun, and sometimes because it makes economic sense for the producer of the good, even if no-one else who will enjoy the good contributes.

Now, there’s nothing wrong with altruism, but people can only afford a certain amount of it, so perhaps we could have even more public goods if we could figure out an efficient way of funding them. How well can we do?

One way of funding public goods is through pledges, as on Indiegogo or Kickstarter. There, a would-be producer of a public good can post a campaign, essentially saying “If people pledge a total of at least \$$x$ by such-and-such a date, I’ll produce this specific public good”. People interested in that particular public good can then pledge money to the campaign, knowing that they’ll be refunded if the total isn’t met. (They also allow campaigns where the money is paid to the campaign even if the target isn’t met, but I’m mainly interested in the kind of campaigns I’ve described above.)

The problem with this is that if someone thinks the total is likely to be met without their pledge, then they have no (non-altruistic) incentive to pledge their own money; they’ll be able to enjoy the eventual public good without contributing. And if they think the total is unlikely to be met even with their pledge, then they still have no incentive to go to the trouble of pledging their money, since the good is unlikely to be produced either way.

It’s only in the unlikely scenario that they think their pledge will be crucial in reaching the total that they have an incentive to make their pledge. So this threshold pledge system, or street performer protocol, as it’s sometimes called, still often relies on a degree of altruism.

One interesting alternative is the so-called Wall Street performer protocol, in which bonds are issued for the completion of a specific public good; they pay out to the owners of the bonds when the public good is supplied, so obviously the owners have an incentive to ensure that the good is supplied. Perhaps I’ll discuss this idea further in a future article, but I have another idea to tell you about in this one.

When a public good can be sufficiently well specified, another option is to offer a bounty to anyone who provides that good. This is sometimes used for fixing bugs in open source software; see, for example, Bountysource. Anyone with the desire can pledge money to increase the bounty for fixing a bug that particularly bugs them.

One problem with these bounties is that they encourage unnecessary duplication of effort. Whoever is the first to provide a solution to the bug collects the bounty. If two or three people are simultaneously working on the same bug, the runners-up aren’t compensated for their effort, even if their eventual solution is better in some way than the first solution provided. This risk of non-compensation might reduce the incentive for talented people to try to collect the bounty.

A particular feature of one obscure bounty website caught my eye, though. The only project currently using BidforFix is OsmAnd. Now, the OsmAnd developers can look at which features people are bidding for at BidforFix and indicate their current top priorities.

This got me to thinking: If someone was particularly interested in OsmAnd’s current third priority, and wanted it done sooner, they would have an incentive to bid for that task in an attempt to raise it to the second, or even the first priority, especially if they believed a small additional bid would be sufficient to raise the priority.

Then, I imagined a situation where someone announces: “I’ll provide either public good X or public good Y. X will take twice as much work as Y, so if, before a week from next Tuesday, the public pledges more than twice as much money for X as they do for Y, I’ll take the money pledged for X and provide X; otherwise I’ll take the money pledged for Y and provide Y.” Then, the public knows exactly when a pledge will be crucial in determining which of the public goods is provided.

Suppose there aren’t any pledges yet, and Alice comes along and reasons “Having Y produced instead of X is worth \$6 to me, but a \$1 pledge is sufficient to tip the scales, so I’ll pledge \$1 for Y”.

Then Bob comes along and reasons “I’d pay \$9 for X to be produced instead of Y, but a \$3 pledge is enough to tip the scales, so I’ll pledge \$3 for X”.

Then Alice comes back and discovers that X is leading, so she pledges an additional dollar for Y.

This continues until Bob has pledged \$9 for X, and Alice has pledged \$5 for Y.

But then along comes Cynthia, who’s willing to pay \$4 for X to be produced instead of Y, so she pledges more money towards X, until Alice has pledged all \$6 for Y, and Bob and Cynthia have pledged a total of \$13 for X.

Of course, more people might come along and pledge more for each of X and Y, but I hope you get the idea.

The effect is that the producer of the public good receives some compensation for their effort, according to how much the public actually wants that good, rather than another. And the important point is this: this happens even without any altruism on the part of the people making the pledges.

It’s not perfect, though; people who would enjoy either X or Y, but are indifferent about which one is produced, still have no non-altruistic incentive to make a pledge, though if they are altruistic, they can safely pledge to both campaigns, knowing that they’ll only have to pay for whichever one wins.

And it’s not only indifferent people who lack incentives to reveal the full value to themselves of the public goods. For example, Y might be worth \$150 to Alice, but because X is worth \$144 to her, she’ll bid no more than \$6 to have Y provided instead of X.

But perhaps this idea will, in some situations, improve on existing mechanisms to non-coercively get the public to pay for public goods.