The City Attorney explains what a measure legally does. The City Auditor should explain what its accountability promises can prove.
Berkeley tax and bond measures routinely promise independent audits. The phrase sounds like a guarantee of accountability. But it tells voters only who will check—not what the auditor can determine.
An audit might trace where the money went, confirm that expenditures were legally permissible, or evaluate whether spending produced the results presented to voters. Those are different levels of assurance, yet ballot arguments collapse them into one reassuring word: audit.
Berkeley should make that word mean something specific. Every proposed ballot measure whose promises depend on future City implementation should receive a short, standardized auditability analysis from the City Auditor.
The ballot process already provides a model. The City Attorney prepares an impartial analysis explaining what a measure legally does. The City Auditor should prepare a parallel analysis explaining the reach and limits of its promised audits. Both would accompany the official ballot materials without recommending how anyone should vote.
The Auditor’s analysis should answer four questions.
Campaigns often name projects, services, and benefits that are not binding commitments. The analysis should distinguish legal requirements from amendable plans, examples of possible spending, and campaign statements that create no obligation.
Voters should know which promises will remain enforceable after Election Day.
If a measure promises results, it should identify the evidence needed to evaluate them: starting conditions, expected quantities or service levels, completion dates, cost assumptions, and outcome measures.
Long-term programs need flexibility. But the City can permit substitutions while requiring disclosure of material changes and an explanation of how the revised plan advances the same voter-approved results.
Without a durable baseline, the government can redefine success after the money has been authorized.
The analysis should state whether the audit will examine:
All three are valuable. But an accounting or compliance audit should not be presented as if it guarantees performance.
For example, a general-tax audit may verify lawful General Fund spending. It cannot prove that a fire station remained open unless the measure requires that result. A bond audit may verify eligible capital spending without establishing that projects emphasized before the election were completed.
These limitations do not dictate a vote. They tell voters what protection the audit language does—and does not—provide.
Meaningful auditing requires authority, reliable data, expertise, staff time, and funding. The analysis should identify who will conduct the audit, how often, what records must be maintained, and whether the responsible office can evaluate performance.
It should also explain what happens after a shortfall is found. A recommendation without a deadline, a responsible party, and public follow-up may document a problem without producing a correction.
The Auditor would not need to invent professional criteria. Public-finance and government-auditing organizations publish authoritative guidance. In a City-commissioned analysis, for example, the Government Finance Officers Association recommended that Berkeley require recurring revenues to equal or exceed recurring expenditures—a recommendation the City Auditor later repeated.
An auditability analysis should identify applicable professional benchmarks and disclose when a measure’s design does not incorporate them. It need not declare the measure unworthy; it should make the limitation visible.
The Auditor should apply a standard template early enough for measure sponsors to correct deficiencies. A final version should appear beside the City Attorney’s impartial analysis.
This would improve measures rather than merely criticize them afterward. Councilmembers would know what must be defined, campaigns could describe audit protections accurately, the Auditor could identify resource needs, and voters could distinguish permissible spending from demonstrated results.
Most importantly, the analysis would remain separate from any personal endorsement. An auditor may believe a tax or bond is urgently needed while still explaining that its public promises are not binding or its results cannot be measured. Professional review should clarify the choice, not make it for the voter.
Berkeley already asks voters to trust independent audits. It should tell them what that trust buys.