Persistent Commitments

Persistent Democracy can be applied very directly whenever we're making a decision we can "take back", or is easily reversible. Choices of public officials, legislative or administrative codes, and even budgets are situations where we can quickly change our decision and aren't locked into any kind of outcome.

Not all decisions have that structure though, and sometimes a decision is permanent in some way. We've already hinted at one way to handle these situations by using persistent prioritization, since that system transparently chooses in what order some time-dependent or consumable supply of a resource should be allocated. However there are some irreversible decisions where it isn't possible or reasonable to simply plot all potential options on the same linear curve as with persistent prioritization. In this chapter I'll describe a system we can use for these more complex decisions, and then in future chapters use this system in particular situations.

Description

Say there's some one-time decision that needs to be made. We can simply use normal persistent quadratic range voting to nominate and select different "decision" documents, each stating what final decision should be made. These decisions can be only partially persistent since eventually a final decision will be made, but the election can retain many of the benefits of persistence by having multiple vote update cycles before the decision is finalized. For example if a decision needs to be made in thirty days vote changes can be published every three days to let the decision find some equilibrium.

The danger we have to mitigate is some kind of strategic action happening right before the decision deadline. We should do whatever we can to encourage the decision to converge to some generally acceptable option without surprising fluxes of incoming voters or options throwing everything off. One simple way is to simply freeze all allocated weights in the final few cycles, essentially fixing the possible options and only allowing different range approvals between them.

Here's a more specific breakdown:

• A one-time decision needs to be made, let's say within thirty days. Any interested citizen can place weights on the entire decision, and weights can be split between nomination or to increase the strength of their range ballots. Each voter only has one positive range ballot and one negative range ballot.
• As usual at any time voters can change the scores on their range ballots or place nomination/anti-nomination weights. On some shorter cycle, such as every three days, all these changes are published.
• In the last few cycles voters can no longer add or remove weights from this decision or nominate new options, thereby fixing the number and relative weight of the final voters and the option set.
• The final decision is made by these voters as they move their weights around and change their range ballots.

Since this structure is making irreversible decisions, it makes most sense to use when making recurring commitments on a regular cycle, since that hopefully means each individual decision is less serious. In some of those situations it could make sense to have some "default" document that will be used if no others are proposed. Future chapters will apply this system to specific problems.

Benefits

• Bridges the gap between persistent and discrete decisions, possibly even offering a transition path for groups to move toward persistent democracy without committing all at once.
• Allows groups to democratically make commitments in a cautious and consensus-oriented way.

Potential Objections

The largest problem with this system is that it partially reintroduces the issues of discrete elections, and we have to do some work to prevent surprising strategic outcomes.

Open Questions

Are there better ways to mitigate strategy? Locking weights and not allowing more nominations for the last few cycles is quite arbitrary.