Abstract
This post raises a procedural concern that I believe the ArbitrumDAO needs to address: the practice of bundling multiple, substantively distinct proposals into a single AIP. I argue this undermines the quality of our governance and should be discouraged or formally prohibited.
The Problem
A recent example illustrates this clearly: ArbOS 60 Elara bundles three separate governance actions into one vote, a new multidimensional gas pricing algorithm, an increase to the Stylus contract size limit, and a two-year delegation of base fee modification rights to Offchain Labs. Each of these carries meaningfully different risk profiles, stakeholder implications, and tradeoffs.
A delegate who strongly supports the Stylus contract size increase but has reservations about granting Offchain Labs broad, fast-acting discretion over the minimum base fee, as MconnectDAO raised in the thread, has no way to express that nuance. They must vote yes or no on the bundle, effectively being coerced into accepting proposals they might otherwise reject in order to support the ones they do.
Why This Matters
Bundling creates a structural incentive to attach controversial or weaker proposals to popular ones. It dilutes accountability, suppresses dissent on individual components, and makes it harder for the community to form clear, informed opinions. It also muddies the governance record when a bundle passes, it’s unclear which parts the DAO actually endorsed.
What I’m Proposing
I’d like the ArbitrumDAO to adopt a norm and ideally a formal guideline that each AIP must contain a single, coherent governance action. Where a proposer believes multiple changes belong together, they should justify that explicitly and allow the community to challenge the bundling before a Snapshot vote proceeds.
Call to Action
I’d welcome feedback from delegates and the broader community. Has bundling benefited governance in cases I’m not considering? Should this be a soft norm or a constitutional requirement?