This post is part of the Unweirding Boards sequence and a close partner to An Outcomes-First Case, in which I began to outline a Theory of Change (ToC) for the good governance of EA organisations.

A Model of Governance

While The Good Governance Project has so far focused on formal governance of EA organisations (through Boards of Trustees), I think it's worth thinking about an expanded model of governance that looks a bit like:

Type of governance	Lever	Description
Board	Oversight	A board with context and decision-making authority oversees and supports the work of an organisation
	Compliance	A board ensures regulatory requirements are met
	Performance Monitoring	A board participates in the setting of org- and individual-level targets, and monitors progress against these
Funder	Performance Monitoring	A funder sets clear targets and monitors progress against these
	Allocating Resources	A funder prioritises cause areas and specific interventions, and assesses individual requests for funding
	Accountability	The EA community engages with organisations (critiquing ToCs, BOTECs etc)
Community	Accountability
Community	Norms	The community sets norms and expectations, expecting that these are upheld by organisations

Based on conversations with ~30 people (founders/execs/board members) since spring 2025, my sense is something like:

What Could Go Wrong?

When I first shared this model, and an observation that the current state was a "a recipe for not good things", @calebp prompted me to share:

a specific fictional story of failure that [I] think is:
realistic (e.g. [I]'d be willing to bet at unfavourable odds that something similar has happened in the last year)
seems very bad (e.g. worth say 25%+ of the org's budget to fix)
is handled well at more mature charities with better governance
stems from things like 'orgs and individuals held to

...