Scaling Mutual Expectations

In order to effectively cooperate with others, a crucial prerequisite step is forming shared mutual expectations. The processes we use to form these expectations are often invisible to us, implicit in our daily lives, but with a little consideration we can unveil them and expose them to explicit intervention in the process of mechanism design.

When we encounter expectation formation most explicitly in our everyday lives, it is most often in the context of dyadic relationships. When the expectations are purely formed between you and one other person, you have the most creative control, so these situations might stand out to you most. For example, dyadic expectations are formed when agreeing when to hold meeting, or what restaurant to go to dinner together, or how much you will pay a handyman in exchange for what services. These mutual expectations are sometimes symmetric -- for example, if we agree to meet at a Thai restaurant, and split the bill 50/50. Mutual expectations may also be asymmetric -- for example, if a handyman fixes a leaking sink, I am unlikely to repay him by driving to his house and fixing a sink of his in return. In each of these examples, ad-hoc negotiated expectations are used, which are most common at this dyadic interpersonal level.

When coordinating with larger groups of people, asymmetric expectations get even more complicated. For example, on a product team, each member will have different expectations of them -- a designer to specify the visual look & feel and basic user interaction patters, some number of engineers (each potentially specialized into different domains), and a project manager to manage priorities and feedback from key stakeholders. In this example, role-based expectations are used. Role-based expectations are most commonly re-used from some common cultural understanding (since negotiating many roles from scratch between many people would be horridly inefficient). These expectations are most common at the project level.

For coordinating with even larger groups, vision-based expectations are more common. In this style of coordination, a common vision is socialized across potential members of the group, and those potential members make the choice to affiliate, and creatively bring their own skills and ideas towards furthering the goal. This style is less structured, but encourages creativity from individuals. In order for anyone to generate a new project, they need to be able to understand the group relative to its goals. The more clear the vision is, the more likely group members are to creatively engage with the vision and spawn new projects. These expectations are most common at the organization or movement level.

Each of these strategies for coming into expectations with one another has strengths and weaknesses, which are generally scale-dependent. Often, one relationship can layer all of these kinds of expectations on top of each other -- working at one company, we have a shared vision; in the context of that shared vision, I have a role on my team; and for any working relationship, we may have specific negotiated mores. This fractal nesting of expectations allows us to garner the benefits of each of the kinds of expectation formation while still remaining tractable at each scale.

	Ad-Hoc	Role-Based	Vision-Based
Natural Scale	Dyadic relationship	Project	Organization or movement
Coordination Complexity	Very high	Medium (per member)	Low (per member)
Asymmetric contributions	Possible when pushed by participants	Asymmetric by default, but asymmetries may not cleanly fit individual capabilities	More likely when members are creative, less likely otherwise
Creative Contribution	Likely	Difficult unless creativity is built into roles	More likely when the vision is strongly felt

As we consider formalizing coordination in mechanism design, we should be aware of which of these expectation formation styles are most relevant. Are we customizing expectations line-by-line, as in ad-hoc negotiated expectations? Are we relying on some established understanding of roles, or crafting new roles which mesh more cleanly with our goals? Are we dependent on some communicated vision, which should be made clear to all participants for their creative engagement? With greater consciousness of these expectation formation styles, we are able to mitigate weaknesses and capitalize strengths to make sure our goals are most effectively implemented in mechanism design.