What Polkadot governs

Polkadot is a layer-0 network: a central relay chain that provides shared security and interoperability to the parachains connected to it. Its native token, DOT, is used to stake for validation, bond parachain slots, and—most relevant here—to govern the protocol itself. The relay chain went live in May 2020, and DOT was redenominated 100× in August 2020.

Unlike most token-DAOs on this directory—which run off-chain token-weighted votes on Snapshot and then execute through a multisig—Polkadot's governance is entirely on-chain. An approved referendum executes its own privileged call directly against the runtime, with no council or foundation as an execution intermediary. That makes it one of the purest large-scale examples of on-chain governance in production.

From Governance V1 to OpenGov

Polkadot's original model (“Gov1”) split power between a token-elected Council, a Technical Committee, and public referenda. In June 2023, Polkadot enacted OpenGov (formerly “Gov2”), which abolished the Council entirely and handed every decision to direct, token-holder referenda.

The key unlock is parallelism: Gov1 could decide roughly one referendum at a time, whereas OpenGov lets many referenda run simultaneously across different classes of decision, each with a risk-calibrated set of rules. A tweak to a treasury tip and a change to the Root runtime no longer compete for the same queue.

Origins and tracks

Every privileged action in OpenGov is dispatched from an Origin, and each origin is bound to a track with its own capacity, deposit, and timing rules. There are 15 origins, ordered roughly by how much damage a bad decision could do:

• Root — the most powerful origin (runtime upgrades, setting the code itself). It accepts only one proposal at a time and carries the longest lead-in and highest deposits.
• Whitelisted Caller — a fast path for changes that the Technical Fellowship has vetted and whitelisted, with shorter timing than Root while still requiring a token-holder referendum.
• Treasury spenders — a graduated ladder from the Small Tipper (up to 250 DOT) through Big Tipper, Small/Medium/Big Spender, up to the Treasurer (up to 10,000,000 DOT). Bigger asks route to higher-scrutiny tracks.

Riskier tracks demand higher turnout and slower timelines; a cheap tip can clear quickly, while a Root change must survive a long, high-bar decision window.

Conviction voting and the referendum lifecycle

OpenGov weighs votes by conviction: a voter multiplies their voting power by voluntarily locking DOT for longer after the vote resolves. The multiplier runs from 0.1× (no lock) up to 6×, where the 6× level locks tokens for 32 lock periods (roughly 224 days). Voting power is tokens × conviction, so someone who is willing to be locked in behind a decision counts for more than a whale who wants their DOT back immediately.

A referendum passes only if two curves are both satisfied through a confirmation period: Approval (the conviction-weighted share of aye votes) and Support (turnout, measured against total possible votes). Both thresholds relax over time—an uncontroversial proposal can meet a high early bar, while a contested one has to wait for the curves to loosen. The lifecycle is: Lead-in → Decision → Confirmation → Enactment, per the OpenGov reference.

DOT holders who don't want to vote on everything can use multirole delegation: delegate voting power to a different trusted account per origin, so you might follow one expert on treasury spends and another on technical upgrades.

The on-chain Treasury

Polkadot runs a large, protocol-owned treasury that spends only by referendum. Per the treasury docs, it is funded by 80% of transaction fees (the other 20% goes to block producers), 15% of annual DOT inflation, and slashed validator/nominator stake. Crucially, unspent funds are partially burned at the end of each spend period—a deliberate pressure to deploy capital rather than hoard it, a problem many idle DAO treasuries never solve.

Spending flows through the Tipper and Spender tracks above. For recurring, curated work, the treasury also funds bounties: a parent bounty reserves DOT under a curator (usually a multisig), and child bounties pay out from it without a fresh referendum for every disbursement—an on-chain answer to the working-group funding problem.

The Technical Fellowship

OpenGov removed the appointed Technical Committee and replaced it with the on-chain Polkadot Technical Fellowship: a permissionless, rank-based collective of the protocol's technical experts. Ranks run from 0 (Member) to 9 (Grand Master), votes inside the Fellowship are weighted by rank, and members are promoted through the Fellowship's own referenda (see the live fellowship dashboard).

The Fellowship's main governance power is whitelisting: it can flag a proposal as technically sound, routing it to the faster Whitelisted Caller origin. It cannot enact anything alone—a whitelisted proposal still faces a full token-holder referendum. This is a clean separation of expertise from authority: technical review speeds a decision up, but token holders still decide.

How Caper approaches this

Conviction voting solves a real problem—letting committed holders outweigh mercenary ones—by asking voters to opt in to a lock. Caper reaches the same goal without a separate locking step. A member's weight on a proposal is computed directly from the stake they already hold in that specific caper: in the contract it is (your vote tokens × your votes) / (vote supply × tokens in circulation), and a timelock blocks last-minute vote-bombing before a tally. Influence and exposure move together by construction—you weigh more because you already have more capital at risk in the outcome, not because you signed up for a longer lock.

See also sibling network DAOs Arbitrum and the Optimism Collective, and the networks directory.