The Ledger and the Life

In 2020, Tracie Ann Grijalva received a conditional offer of employment. A background report disclosed her continuing exclusion from federally funded health-care programs and the 2011 revocation of her certified nursing assistant license that had occasioned it. The offer was withdrawn. In 2025, the Ninth Circuit held that the continuing exclusion remained reportable, but the old revocation did not. The court nevertheless affirmed judgment for the reporting company: its interpretation of the law had not been objectively unreasonable, and the violation was neither negligent nor willful.¹

The distinction she established did not give her a victory in the judgment. It changed the account of what a report could properly contain. We should be able to recognize the importance of that change without describing the two achievements as though they were the same.

A legal order has good reasons to distinguish an unlawful act from the conditions under which its author must pay for it. Reasonable interpretation matters; so does the possibility of acting under law whose application has not already been decided. To remove every protection against liability would not abolish uncertainty. It would make the consequences of uncertainty fall differently, often on people attempting something for which no settled answer is yet available. But a doctrine which protects an actor against an unexpected liability leaves someone else living with the result. The law's reason for withholding compensation does not make the loss imaginary.

An institution can hear an objection, examine the evidence, distinguish propositions previously run together, and give a reasoned answer which leaves the person who brought the objection without relief. We cannot judge the arrangement solely by whether that person wins. We cannot judge it solely by whether the arrangement has completed its procedure either. What was made to happen, what can now be changed, and what must still be borne are questions that survive the most orderly disposition of a file.

The books have arrived here by following things that leave one set of hands and acquire consequence in another: a sealed claim, an instrument of credit, an output accepted as grounds for an act. Much of civilization consists in making these passages ordinary. A person can rely on work performed elsewhere, commit resources before knowing every future use, or obtain help from someone whose competence he cannot reproduce. The achievement releases time. Without it, nearly every undertaking would begin with an impossible inquiry into the means by which it had become possible.

Computation enlarges that achievement. It can find an inconsistency that a practiced eye has passed over, preserve a version which would otherwise disappear, and make a check available to someone who once had to purchase another person's assurance. A capable assistant can give an unfamiliar question a usable form. An institution can discover that its confident decisions have been resting on a distinction its own records failed to preserve. These are reasons to build better systems and to use them with ambition. The constitutional demand grows with their usefulness, because usefulness gives institutions more occasions to entrust them with acts that other people will have to live under.

Consider what happens when an institution describes such an act as complete. The calculation has returned, the payment has settled, the restriction has entered the account. For the process that performed it, there may be nothing left to do. Elsewhere, work is beginning. Someone has to arrange the money that did not arrive, establish why an entry is wrong, or learn which of several offices possesses the authority to reconsider it. The system's completion becomes another person's task. If the time spent on that task does not appear in the institution's account of efficiency, the saving will look larger than it is.

A timely decision can prevent the very injury a slow institution would have caused. Decisions need not wait while all imaginable objections are assembled. The institution must make provision for the work its decisions will require, including the work occasioned by error. An automated demand should not be made cheap by obliging its recipient to reconstruct the evidence that should have supported it. A review should not consume the period in which the disputed interest could still be protected. Where a rapid act is necessary, authority for that act must include responsibility for what follows it.

The continuing institution matters more as its practical actors become easier to replace. A model is updated, a vendor changes, an invocation ends. The person now asked to explain the act may have had no part in making it. We know this difficulty from older offices: the successor inherits papers and powers before acquiring a memory of the decisions made through them. Continuity was never only a fact about the lifespan of an official. It was something an institution had to arrange.²

That arrangement should carry obligations with the powers. An institution cannot invoke its continuing identity to collect what is owed to it and become a succession of unrelated moments when asked what it owes. The officer who receives the complaint need not impersonate the vanished decision-maker or defend every choice of a predecessor. She needs access to the evidence, authority to investigate, and a route by which the institution can correct what it has made consequential. Inheriting responsibility should include the power to depart from the inherited answer. Otherwise continuity becomes the organized repetition of a decision nobody present is allowed to reconsider.

Some of this work can itself be automated. A corrected premise can be carried into the later decisions that used it; an expired authority can cease to authorize; an agreed consequence can follow without a second petition. The more reliably these tasks are performed, the less of a person's life has to be spent seeking someone willing to perform them. The difficult cases will remain, but routine protection should not require an exceptional official. A republic that depends on finding the one considerate person behind a counter has made consideration do the work of a right.

There is an attraction in being that official, and another in never having to be. A rule can spare its administrator the discomfort of disappointing someone personally. A recommendation can supply reasons that were formed without the administrator having to expose a judgment of her own. Often those reasons are better than the ones she would have supplied. Agreement may be exactly right. The relief becomes dangerous when following the recommendation is treated as sufficient evidence of responsibility, while departing from it means assuming every risk of what happens next.

An institution organized on those terms will possess an override in its procedures long after it has made an override professionally unreasonable. The conscientious official can see what a refusal costs the applicant and what a grant might cost herself. One cost is already present; the other may someday be attributed to her by people who will know the outcome before examining the decision. Under that expectation, the safest recipient of relief is the person whose future needs the least indulgence. The apparent caution selects among lives.

We should require the official to answer for her authority, evidence, reasons, and treatment of comparable cases. We should not require her to warrant another person's future. A justified release may be followed by harm, just as a justified refusal may deprive someone of a possibility that would have gone well. Later evidence can expose negligence or favoritism; the outcome alone cannot tell us whether the earlier judgment was defensible. Institutions need the courage to preserve this distinction after an outcome has made it unpopular. Where the distinction is abandoned, permission to exercise judgment can become an instruction to choose the answer least difficult to defend afterward.³

An institution's obligations cannot be settled by a verdict about fault alone. Blame, repair, compensation, and the prevention of another injury need not depend on identical findings, and a serious institution must say which question it has answered when it declines to act. Someone may have behaved reasonably while an arrangement still requires alteration. A rule may have been faithfully applied while the allocation it produces can no longer be defended. The fact that no individual deserves condemnation does not establish that everyone else should continue to bear the cost.

There is nothing ennobling about making a person perform deliberation that a competent system could have spared him. People have reasons to spend their attention elsewhere, including reasons a defender of participation ought to respect. A life need not become an unending inspection of the arrangements that sustain it. To place every responsibility for vigilance on the person exposed to power would reproduce, in the language of freedom, the burden the books have been trying to move.

The distinction is between accepting assistance in one's own life and authorizing an institution to act in the lives of others. A person may follow good advice for years. An institution which adopts advice as grounds for an exclusion owns that adoption, even if the advice was excellent and nobody could have produced a better forecast. Accuracy can help justify the choice. It does not supply jurisdiction. A system which serves its principal faithfully may still be serving an end the principal has no right to impose on everyone affected by it.

The political standing of conscious life does not depend on a settled account of how consciousness arose. People do not acquire their standing by defeating a machine at the activity on which an institution happens to place a price. The loss of an economic advantage would not establish that their lives had lost their claim upon the order they inhabit. We owe a defense of the purposes for which power is exercised, including to those whose contribution an improved production process no longer needs to purchase.

Such a defense cannot be given once and inherited without further work. An arrangement acquires uses its designers did not foresee, beneficiaries who did not build it, and people subject to it who never had an occasion to consent. A record is especially capable of this later life. It can be copied into another system without visibly changing, although nearly everything that determines the propriety of using it has changed around it. The receiving office sees continuity in the information. The person encounters a new exercise of power.

Preserving evidence of that exercise and limiting the later use of the person's history are compatible duties, though they will sometimes be expensive to perform together. The institution needs enough of the past to answer for what it did. It does not follow that every recipient of its data needs the same past in order to decide what to do next. A later use requires an account of its own purpose and authority. The existence of the record does not furnish that account, and requiring the person to disprove the record will not produce it.

Nor should a person have to establish that he has become new before an institution observes the limits on an old claim. Evidence of rehabilitation can matter profoundly where present judgment calls for it. A reporting limit or a discharge does different work: it limits an institutional power under conditions a person can meet without persuading an official to admire the life subsequently lived. Someone may be uninteresting, unrepentant, or difficult to like and still possess a right. A constitutional limit that becomes available only after a moving account of transformation has been given is already drifting toward favor.

Other claims may remain. Ending a public consequence cannot make the wronged party forgive, and preserving a future for one person does not entitle an institution to destroy the evidence another needs. These differences resist the comfort of a single word. Release is not accomplished by making the past universally unavailable. Accountability is not accomplished by making it universally active. The work lies in deciding what this institution may continue to do, to whom it must answer, and which parts of the history another encounter is entitled to receive.⁴

The task has no final administrator. Institutions can lose their independence, inherit the interests they were meant to inspect, or mistake the continued production of records for continued attention to what those records reveal. The answer is not to require every citizen to become an auditor. It is to sustain enough independent sources of evidence, competence, and authority that an institution's account of itself need not be the last account available. That requires paid work, ordinary offices, and people with time to learn what an apparently settled procedure has been doing. It requires the resources of a public world.

At the beginning of the trilogy, Kakudmī waits through a performance before asking the question for which he has traveled. When the audience is over, the world to which his question belonged is gone. The extravagance of the interval makes the story memorable. Its smaller cruelty is familiar: an answer can arrive intact after the possibility of using it has disappeared.⁵

We cannot arrange that every answer comes in time or that every justified decision leaves a life unharmed. We can refuse to treat the time consumed by an institution as though it belonged to nobody. We can give it duties that begin before a complaint, powers sufficient to correct an act, and obligations that survive the succession of people and processes through which it acts. And we can refuse the opposite continuity, in which a person remains available forever to every inference that can be drawn from a preserved event.

These choices will seldom announce themselves as the founding of an order. They will concern the retention of a version, the terms on which a service may be relied upon, the budget of a reviewing office, or the decision to stop requiring a fact that a form has always requested. Their smallness is deceptive because each will be repeated. What becomes ordinary in those encounters determines how much of a life must be spent asking permission to proceed.

Grijalva's case leaves a reporting boundary established and liability denied. It is an awkward result for anyone looking for a vindication, and a useful warning against making the clarity of an opinion stand for what its claimant received. The next institution that uses the distinction has work left to do. It has to alter the report, the permissible fields, the practice through which the old information reaches a later employer. A proposition about limited authority has to become a limit in an ordinary transaction.

The books end there, before the next report is sent. Someone will decide what it contains. The applicant should not have to live the whole argument again.

Source notes

Footnotes

1. Grijalva v. ADP Screening and Selection Services, Inc., No. 24-2984 (9th Cir., August 15, 2025), opinion, pp. 5–7, 14–20. Judge Clifton disagreed with the reporting distinction, pp. 21–25. This discussion concerns that opinion, not a claim about subsequent employment or litigation. ↩

2. Institutional continuity is an antecedent, not a novelty claimed here. See Ernst H. Kantorowicz, The King's Two Bodies (Princeton University Press, 1957), especially chapters VI–VII. The duty of an institution to inherit answerability with its power is this epilogue's normative application, continuous with the Codex's surviving principal and Volume III's treatment of succession and delegated authority. ↩

3. The distinction between an answerable judgment and a warranty of the recipient's future is developed in “Homo Arbiter,” “The Slippery Right,” and “Against the Protocol Republic,” “Mercy Under Observation.” The official described here is an institutional possibility, not a reported individual. No empirical frequency or universal liability rule is asserted. ↩

4. Paul Ricoeur, Memory, History, Forgetting, translated by Kathleen Blamey and David Pellauer (University of Chicago Press, 2004), epilogue, “Difficult Forgiveness,” supplies philosophical ancestry for remembrance without permanent identification of agent and act. The distinctions among purpose-specific public release, clemency, and a wronged party's forgiveness are argued in “Homo Arbiter” and Appendix D. They do not follow as legal rules from Ricoeur. ↩

5. Śrīmad-Bhāgavatam, Canto 9, chapter 3, verses 29–32, translated by A. C. Bhaktivedanta Swami Prabhupāda (Bhaktivedanta Book Trust, 1977), reproduced at the opening of the Codex. The closing application concerns delay; the story is not evidence for a claim about computational speed. ↩