Other allocation methods

In addition to the simpler allocation methods, there are also a range of more complicated allocation methods. I'm not going to pretend here that I am an expert in any or all of them, nor that what is depicted here is sufficiently robust for a wide variety of purposes, but I do want to include some examples to illustrate what sort of options and what type of thinking could be used in this space.

Bidding with time

In an exchange economy people can bid with money, and this sort of bidding can result in something called "price discovery", where the bids begin to hone in on the "true" price of the asset. This is done regularly with houses through auctions.

But in a giftmoot economy, bidding with money isn't an option. However, there is an obvious resource that everyone can use to illustrate their preferences: time. So instead of bidding with money, people could instead bid with time. I'll give two examples, but I imagine that there are many more variations of this.

In the first example, people bid on the latest time they are happy to receive the product. On the time they bid they for they are given the highest priority, and after that time they leave the bidding pool. If there are too many bidders for a timeslot, the products are given to a random subset of the most timely bidders. If there are too many products for bidders, then the next timeslot's bidders receive their product early.

For example, imagine a company that produces 10 products per day, and opens the bidding the day before the first product will be shipped. If a person were to bid for day 1, they would receive the highest priority for day 1. If 11 people bid for day 1, 10 will receive products, and the 11th will be removed from the bidding pool. However, if only 8 people bid for day 1, 2 products are shipped to later bidders early (for example, day 2 bidders). This motivates people to bid for the latest timeslot that they can accept, given that a bid for an earlier timeslot might exclude them, but a bid for a later timeslot could mean they receive the product early.

Bidding on allocations

Another type of bidding could occur for specific blocks of a resource. Imagine people who want to bid on dates that they could go to a holiday house. They are not necessarily trying to bid for the earliest dates, but the dates that suit them the most (maybe their birthday or anniversary, when they get a break from work, or to line up with favourable weather). Obviously time is not a relevant resource for the bidders, so in this case it might be useful to grant each bidder a set of bidding points. This means that bidders are placed on somewhat equal footing when the bidding starts, rather than wealthier people starting with an advantage.

Let's say a bidder is given 100 bidding points, and they can distribute them across the dates they desire in any way that they want. They could allocate 1 point to each date, or put all their points on their favourite date, or distribute them in a bell curve around their favourite date. Once all the bidders have allocated their points, each individual date is resolved in order of highest bids to lowest bids, with the losers of each bid having their losing points transferred to their next highest bid. (You can also do this backwards, starting with the lowest winning bids and working up.) For example, if person A bid 100 points for the 1st of January, and person B bid 80 points, person A would win the date and person B would have their 80 points transferred to their next favourite day (perhaps the 2nd of January). If a person is out of bids, their points disappear. This ensures that a person is likely to get a date of their preference, but also ends up favouring those people who spread out their bids. A person who puts all their points on one date could find that another person has also put all their points on the same date - leaving one to receive the date, and the other to have their bid closed because their points cannot be transferred.

In this system, people could also bid to win multiple dates, or multiple allotments on multiple dates, which means that it could be used for shipping allotments, allowing different parties to bid for different amounts of resources in the same auction.

Cake-cutting

Cake-cutting is the name of an area of study in maths where resources get divided up using some procedure that satisfies all of the participants. The example resource is often cake. What's interesting about cake-cutting procedures is that the resource doesn't have to be uniform. For example, a piece of cake could have a portion with blueberries and a portion with raspberries, and the participants could each value the fruits differently. A more real-world context could be land that needs to be divided, where the land differs in characteristics from place to place.

Cake-cutting procedures can also be made so that everyone is happy, that everyone believes they got the value that they desired, or so that everyone believes they have received more than the value they desired.

The simplest cake-cutting procedure is one that works for two people, called "cut and choose". In this system, one person decides how to cut the cake into two pieces, and the other person chooses which piece to have. The first person is therefore motivated to make the cut as fair as possible, because if the cake were divided into unequal portions, there is a chance that they would receive the less valuable portion. (This isn't necessarily a problem if they are not motivated to get half the value of the cake, but just some minimal value to satisfy themselves.)

The more people are involved, the more complicated cake-cutting procedures get, until the number of steps required (the number of cuts and the number of choices) is ridiculously high, and all the participants would be long dead before the procedure could be completed. However, for smaller groups of people, the number of steps can sometimes be more reasonable.

Democratic cake-cutting

Another way to perform cake-cutting might be what I will call "democratic cake-cutting". In this version, any number of participants are split into two groups, and each group is asked where they would cut the cake. Each group needs to collectively deliberate to come up with an answer. Once both have proposed a cut, one group is chosen at random to cut the cake, and the other group then gets to choose.

Once each group has received their cake portion, the process could begin again, with each group being split into two smaller groups that need to deliberate on cuts, and so forth. For 100 people, this would be somewhere in the vicinity of seven stages of cutting. However, the point is not to continuously go through stages of cutting. Instead, because each group has had to deliberate on what would be a satisfactory portion of the cake, and because there is the option of continuous cutting, the participants will choose a cut that sufficiently provides for everyone in the group (because no member can be sure otherwise that they would not be the person to miss out). Therefore, after one cut, the participants should be able to distribute the resources according to their previous deliberation.

It's unlikely that this particular method would be as mathematically fair as the more rigorous cake-cutting procedures, but it would likely be quicker and it would engage with deliberative qualities that mathematical cake-cutting does not encounter.