Experimenting with Random Sampling in the Citizens’ House
What is random sampling?
Random sampling is a widely used statistical method in which all members of a population (all Citizens) have the same probability of being selected. Random sampling does not guarantee that a particular sample is a perfect representation of the population, but rather allows for valid conclusions to be drawn about the entire population based on the sample. Another way of saying this is that the random sample approximates the full population. This is due to the equal probability of selection.
This method is widely used in a variety of disciplines when little is known about the entire population and/or extracting information from the entire population would be costly or infeasible. This is a powerful tool when considering the expansion of Citizenship and the increase in the number of rounds per year.
Does random sampling presuppose Citizens are interchangable? - joanbp
It doesn’t! In fact, it makes no assumptions about Citizens at all, which is why one of the major advantages of random sampling is the reduction of bias that can occur via subsets selected in any other way. Opt-in methods are particularly susceptible to various biases and can create opportunities for capture by a small, active, and likely unrepresentative subset. Random sampling only assumes that each participant had an equal probability of selection.
Why use random sampling now?
Some of the most common feedback in Round 3 was the need to more effectively scale the Retro Funding process (it was too time consuming for badgeholders!) and to provide more predictability for builders. There are a variety of measures we are introducing to address this feedback, including more frequent and narrowly scoped rounds. When scaling to four rounds per year (a 4x increase), we felt it was important to keep the expectation on badgeholder time relatively fixed (or even to reduce it.) Allowing badgeholders to opt-in to rounds limits the number of voices that are heard and may skew or bias the results towards a small subset of Citizens that are able to dedicate the most time. Full participation in all rounds, on the other hand, demands too much time and effort of the whole group. By requiring full badgeholder participation in two rounds and splitting participation between the other two rounds using random sampling, we believe we can achieve this goal. Throughout Season 6, we plan to experiment with random sampling as an alternate or additional method to reduce the workload required of any one badgeholder. More on why we believe this is a worthwhile experiment below.
Why random sampling, specifically in the Citizens’ House?
Should only the most highly engaged Citizens be selected to participate? - gonna
The Token House scales via political representation, which while increasing participation and (potentially) context, also concentrates power. Delegation increases participation but concentrates power among large delegates. In this context, elected Councils, which delegate decision making power over specific decisions to a one member/one vote Council, can actually reduce the concentration of power in the Token House. A weakness of this model, which we will have to actively combat, is that it’s easy for the ~50 most highly engaged delegates to become responsible for making most decisions.
In contrast, the Citizens’ House operates via a one Citizen, one vote model which distributes power evenly among the Citizenry, appearing much closer to a democracy. Elections, delegation, and/or other political scaling methods will concentrate power in the Citizens’ House. The downsides of these approaches in traditional democracies are thoroughly outlined in Open Democracy. One of Hélène Landemore’s main hypotheses is that random sampling (sortition) can be used to avoid these downsides without a meaningful impact on collective outputs (paraphrased.)
Reducing the concentration of power is one of the main goals of the Citizens’ House and random sampling is a well tested method that could allow the Citizens’ House to avoid a common failure mode of other governance systems: capture by a small subset or minority interest. Accordingly, we believe it is worth experimenting with random sampling to enable the Citizens’ House to scale by using statistical, rather than political, methods. With power evenly distributed among Citizens, the Citizens’ House may serve as a more effective check on the Token House.
In short, the intention is not that sampling be representative of a smaller subset of a certain type of Citizens, but rather to approximate a response from the entire Citizenry. High context, highly engaged Citizens should absolutely be recognized, rewarded, and respected within the Collective. Random sampling is not at odds with any of those objectives though!
