Disbursement
Search
K

Decision approaches

Comparing the different decision approaches that could be used for selecting priorities, ideas and contributors
A high level overview of the key decisions that are made during disbursement are that voters are tasked with digesting any self selecting any relevant knowledge about an ecosystem before they then move onto selecting the most important priorities, ideas and contributors that could help with generating impactful outcomes for the ecosystem.
  1. 1.
    Knowledge Selection - Available knowledge within an ecosystem will play an important part in all of the selection processes involved in disbursement. Both automated systems and one off pieces of research or analysis can help with creating useful sources of information for voters. Voters having more access to quality knowledge sources will help ensure that they are more well informed when making any disbursement decisions.
  2. 2.
    Priority Selection - Priorities help to highlight the most important problems and opportunities that exist in an ecosystem. Voters will apply their own understanding of the ecosystem and available knowledge to determine which priorities are the most important for the ecosystem at that moment in time. The selection of priorities can often influence which ideas and contributors get selected as the most promising and suitable for addressing those priorities.
  3. 3.
    Idea Selection - Ideas help to articulate a solution which could be executed that would help with addressing one or multiple priorities. Voters could participate in selecting the most promising ideas that can help with addressing the existing priorities in an ecosystem.
  4. 4.
    Contributor Selection - Contributors offer their time and skills to help with the execution of ideas. Voters could help with selecting the contributors that should be compensated and responsible for the execution of the most promising ideas.
How different decisions are handled and composed together will greatly influence the flexibility, scalability and complexity of disbursement decision making.

Approaches for handling priority, idea and contributor decisions

The selection of knowledge is often a personal choice where the voter will self determine what pieces of information they believe are most important, this is naturally influenced by their expertise and background. An open and community wide decision process could be used to select knowledge, this is something worth considering in the future. For this initial decision comparison we will just focus on the selection of priorities, ideas and contributors.
A number of decision approaches could be used to select which priorities, ideas and contributors are the most important or promising. A selection of these approaches include:
  • Fully combined decision - The selection of priorities, ideas and contributors all happens in the same selection process.
  • Idea & contributor combined decision - The selection of priorities happens independently and then the selection of ideas and contributors happens in the same process.
  • Independent decisions - The selection of priorities, ideas and contributors happen all in separate processes. Voters are encouraged to participate in all of the decisions.
  • Passive idea selection - The selection of priorities, ideas and contributors happen as separate decisions. Voters are encouraged to participate in idea selection in the areas they are interested in participating.
  • Delegated idea selection - The selection of priorities, ideas and contributors happen as separate decisions. Voters are encouraged to participate in sharing their preferences of which ideas are most promising in a community vote, however the responsibility of actually selecting which ideas will be executed is delegated to the contributors that the voters have selected.
To compare these decision approaches a number of factors have been considered and then applied to each approach to try and determine any strengths and weaknesses of each one.
The most expressive, flexible and least complex of the decision approaches outlined above were those that adopted independent processes for decision making to select the most important priorities, ideas and contributors.
The differences between passive idea selection and delegated idea selection have been explored in more detail here:
Decision approaches analysis

Considerations for handling knowledge

At the point of making any disbursement decision the voters involved will have some pre-existing knowledge based on their own experiences and skill sets as well as having access to any publicly available knowledge about the ecosystem, other ecosystems and relevant topic areas.
Voters will decide how much time they want to spend on identifying new knowledge that could help them with making more informed disbursement decisions. Voters may decide not to digest new information for a number of reasons. One could be because they don't trust the source of the new information. Another reason could be that they believe they already have enough knowledge to make informed decisions. Another reason could be they simply don't have enough time available to go through the knowledge sources.
A disbursement process will benefit from making it easy to access knowledge that is relevant and accurate so that voters are able to make more well informed disbursement decisions. Some example areas that could help voters are if they could easily find out what has been funded historically, what priorities or ideas have been completed and which ideas or contributors have been historically the most effective at generating impact for the ecosystem.
For more complex disbursement processes there could be value in exploring how a knowledge selection process could be introduced before voters select priorities, ideas and contributors. Voters may benefit from being able to share and vote on the relevance and accuracy of different sources of knowledge to help with the process of making the community more well informed for upcoming decisions. The most important pieces of knowledge could then become more well known across an ecosystem due to this collective effort to identify quality information sources.

Prioritising certain disbursement outcomes

A key question to ask when adopting any of these different disbursement decision approaches is what types of outcomes are most important? What outcomes should the treasury disbursement process be optimising for? There could be many different outcomes that are important to each ecosystem, each of which those ecosystems may then want to quantify and optimise over time. The following are simply examples, and not even necessarily good ones! Many other potential outcomes can exist that could be optimised for.
Maximising decision involvement
One potentially desirable outcome for an ecosystem could be that a treasury system maximises the involvement of voters in every disbursement decision. Achieving this outcome would mean the entire community will need to be able to vote on any part of the disbursement process and the systems and processes should help with actively encouraging voters to get involved in all available decisions. The independent processes approach could be one suitable way to achieve this type of outcome as voters would be encouraged to vote on everything and voters would also be able to vote on exactly what they want in each process without any compromise on their preferences. The reasons this property might not be desirable are that it might not be practical to try and expect or encourage voters to participate in every decision in the disbursement process. The larger the ecosystem is the harder this would be to achieve. This higher complexity at scale is a reason why approaches such as passive or delegated idea selection can be considered as alternatives.
Maximising participation scalability
Maximising the amount of participation in the disbursement process would mean making it easier and faster for voters to participate in the disbursement process. The easier and quicker it is for voters to participate in the disbursement process the more easy it will be for the disbursement process to scale the amount of participation to a larger audience. Optimising for this outcome will mean reducing the total time required and overall decision complexity as much as possible to achieve a higher amount of participation across a larger community. If this was the primary concern for a treasury system then considering the delegated idea selection approach could potentially be a suitable option. This reduces some of the required involvement from the voter and leans towards increasing the simplicity and speed of the disbursement decisions.
Maximising both participation scalability and decision involvement
Trying to get all community members to decide on every decision can be a difficult outcome when trying to scale the amount of participation to a larger audience. Time consuming disbursement decisions can easily lead to voter apathy and disengagement, a problem that can increase as an ecosystem tries to scale. One way an ecosystem could achieve a balance in achieving both these outcomes could be by adopting a passive idea selection process. Under this approach all voters could still participate in the different disbursement decisions however at a system and process level it is the priorities and contributor selection that are encouraged as the main decisions that should be made by the community. This then leaves the idea selection as optional, helping to reduce the complexity and time taken to participate which helps with increasing participation scalability whilst also allowing for full decision involvement.