Some laws contain a date on which they cease to operate unless renewed. That expiry provision changes the behaviour of everyone involved long before the date arrives.
Expiry converts inaction into repeal
Ordinarily a law continues until a majority agrees to change it, so inertia protects the status quo. A sunset clause reverses that default entirely.
Renewal now requires an affirmative majority, which means a determined minority able to block legislative action can end the policy without ever winning a repeal vote.
That shift in the burden of action is the whole point of the device, and it is why sunset clauses are often the price of passage.
They make bargaining possible
Legislators who dislike a proposal will often accept a temporary version of it, because the commitment is smaller and a further vote is guaranteed rather than merely promised.
Sponsors accept the limit for a different reason. A temporary law that exists is worth more than a permanent one that does not, and programs in operation acquire defenders.
Both sides are therefore betting on the renewal vote, and each believes conditions at that point will favour them. The expiry date becomes a scheduled rematch.
They distort how costs are presented
Budget scoring usually covers a fixed forward window. A provision that expires inside that window shows a lower total cost than an identical permanent one.
This creates an incentive to attach expiry dates purely for accounting reasons, with everyone involved expecting renewal to be routine when the date arrives.
The scored figure and the realistic figure then diverge, which is one reason long-run fiscal projections and headline cost estimates disagree so often.
Renewal debates are rarely a clean review
In theory expiry forces a genuine reassessment of whether the policy worked. In practice the renewal vote arrives alongside unrelated deadlines and becomes leverage inside a much larger negotiation.
Programs with organised constituencies tend to be renewed with minimal scrutiny, because someone is available to make the case. Those without visible beneficiaries can lapse for reasons unconnected to performance.
Evaluation evidence, where it exists at all, usually arrives too late to inform the vote. The decision is made on the political calendar rather than the research calendar.
Where the device works best
Sunsets are most effective where the underlying facts genuinely change — emergency authorities, temporary tax provisions, and rules tied to fast-moving technology.
They work least well where the policy requires long-term investment, because uncertainty about renewal deters exactly the commitments the policy was meant to encourage.
Designers who understand that distinction attach expiry dates to authorities rather than to programs that must build capacity over decades.