The first principle to note is that a decision is never a state or an atomic point action, but a process. Even if we alone struggle with ourselves.
You can see that two phases always follow each other here: First, alternatives are generated and then those alternatives are “cut out” (lat. decidere: to cut off, to fall out) that are considered inferior.
We call the first phase “proposal phase “ because with DAD we focus on group decisions and the possible alternatives are proposed by the individual group members. Nevertheless, the generation of alternatives also takes place in oE and kE, just not necessarily and exclusively through proposals, but through research. One could therefore say more generally that it is the “phase of generating alternatives”, which is divided into the “research phase” for individual decisions and the “proposal phase” for group decisions.
In the proposal phase, interaction plays an important role. As a rule, moderators are involved in important decisions. They ensure that diverse perspectives on the issue are generated and heard. You can book such moderations with us. Or the formulation of the question as such - it already creates an image in the minds of the participants. This is called framing. You can also get advice on this.
We call the second phase “Voting phase “. Here the evaluation of the alternatives by each individual takes place, so that a result can be calculated for each alternative after all votes have been cast, and from this in turn an overall result.
When we talk about “procedures” in the following, we are talking about individual practices that are applied in these phases.
In the literature, among the decision-making procedures, the “process modalities” (sometimes also pejoratively “process formalities”) are rarely highlighted, such as the way of voting (classic: secret/anonymous ballot). Mostly, it is about the evaluation of the individual alternatives and the algorithm for evaluating the votes cast. An overview of this can be found on Wikipedia under the term “electoral system” [Wikipedia2023-2][Wikipedia2023-3].
We will discuss this in the following.
First of all, a distinction can be made between majority voting, rank voting and score voting (also called “range voting”). This distinction is enormously important. It has game-theoretical, social-scientific and mathematical backgrounds. Because Arrow’s impossibility theorem applies to majority and ranked choice voting, which states that one cannot reasonably reach a joint decision here if one insists on compliance with certain basic requirements. (See [Wikipedia-1:Properties] or also [Arrow1963]/[Wikipedia-4] and [BambergCoenenbergKrapp2019:216ff])
And this unfortunately affects most electoral processes in Western democracies. See [Wikipedia-4:Examples] and also [ElectionScience], [FairVote], [ElectoralReformSociety] and [Poundstone2008].
Warren D. Smith has used Baysian Regret to examine how different voting procedures are perceived in terms of outcome [SmithWeb:BayRegExec]. In this picture, also found in Pounstone’s book [Poundstone2008], the results for a particular scenario are summarised:
It can be clearly seen that with sincere (rather than tactical) voting behaviour, Range Voting performs best. “Appoval” is the simplest of all scoring methods. “Borda”, “Condorcet” and “Instand Runoff” are ranking procedures. “Plurality” is majority voting; the simplest of the ranking procedures.
One can summarise this basic problem as follows: If a person is only allowed to vote for one alternative (i.e. majority voting, which is common in democracies), he or she in effect gives no evaluation on the other alternatives. Majority voting is, so to speak, an extreme form of ranking procedures, because one exclusively determines the first rank. Ranking procedures, which allow all alternatives to be ranked, do say something about all alternatives because they have all been ranked, but it is neither apparent how strongly the preferences differ, nor where the pain threshold lies, so to speak. Does one like the alternative on rank 2 only half as much as the one on rank 1 or are they close behind each other? And: does someone like an alternative that he or she has chosen as the third of eight only less than the first two or not at all? Because the differences in approval/opposition are not visible, the ranking procedures are ordinal procedures, they only measure ordinally. In the ranking methods, on the other hand, the strength of the opposition or approval is explicitly stated for each alternative. The measurement is therefore cardinal.
DAD is primarily aimed at organisations that are free to choose their voting procedures ;-)
And DAD wants to do better.
Therefore, DAD uses the evaluation procedures explained below.
The procedure used as a quasi-standard in sociocracy is an evaluation procedure that looks exclusively at consenT and thus only measures resistance. The acceptance range here is a “mild form of resistance”; resistance already begins at a distance from absolute preference. Several scales are possible here. Common is the scale 0,-1,-2.
In the same way, a scale from 0 to -10 is also common. -10 then stands for “serious objection”.
When every participant has cast his or her vote, the average is calculated for each alternative and the alternatives are then sorted in descending order.
In sociocracy, it is common to make the decision-making process synchronous, i.e. in real time in a meeting, and “serious objections” must be taken into account. Through this “integration of objections” directly in the meeting, new proposals emerge and those with “serious objections” are no longer a choice. In an asynchronous process - as facilitated by Module 2 of DAD - this is taken into account by activating the individual veto. This results in an alternative that receives a veto (i.e. a rating of an individual with the veto value, -2 or -10 depending on the scale) being taken out completely in the subsequent processing and the votes that fall on it are not counted.
Ideally, the veto giver drafts a new, “integrating” proposal already in the proposal phase and comments accordingly. If this is done, the vetoed proposal can be deleted already in the proposal phase. In the app, however, the option Single veto does not technically take into account the behaviour of the veto giver. It merely generates the algorithm intervention described above.
This option should be viewed critically, as it can prevent change and presupposes that the group members who decide understand all alternatives to a sufficient degree and can therefore judge them at all and - see ConsenS vs. ConsenT - that at least their tolerance ranges all overlap, which should rarely be the case for very large groups.
This evaluation procedure actively distinguishes between the acceptance and tolerance ranges and includes the whole spectrum of opposition, abstention and approval. It measures the degree of approval (the preference differences in the acceptance range) and the degree of opposition (the preference differences in the tolerance range). Abstention corresponds to the acceptance threshold. That is, the point at which an alternative is no longer really endorsed, but there is also nothing to oppose it.
The active distinction between approval and resistance now offers the possibility to balance between “suffering” (what one has to endure) and “satisfaction” (what one prefers).
Usually, more weight is given to avoiding suffering than to generating satisfaction. Therefore, resistance can be weighted here. This factor represents a kind of exchange rate between the positive and negative ratings, which probably varies with groups and issues. It should at least be greater than 1 to put resistance above approval. If it becomes too large, then a single veto is again possible.