Public Safety, Accountable Policing, and Civil Remedy

The public contract

The first question in policing is not whether one trusts police in the abstract. It is what a free society needs authorized officers to do, what powers those duties require, and what remedy a person receives when those powers are abused.

The Innovation Party starts from two duties that political factions too often separate. Government must protect people from violence, coercion, predatory crime, and immediate danger. Government must also protect people from unlawful violence, coercion, and humiliation committed under color of law. One duty cannot cancel the other. A victim waiting for an emergency response and a driver subjected to gratuitous force are both owed competent government.

Police therefore need real operational capacity. A dispatcher cannot send an officer who does not exist. A patrol officer cannot investigate a homicide while holding a hospital post, answering the next priority call, and completing evidence paperwork. A detective cannot produce a reliable case without lab capacity, witness support, functioning records, and prosecutors able to review it. Police should not be the default answer to homelessness, untreated psychosis, school discipline, or every family conflict. Yet alternative response systems must be staffed, reachable, and able to take custody of a call. Renaming an obligation does not meet it.

The same realism applies to accountability. A policy manual does not restrain force if training contradicts it, supervisors ignore it, cameras are not activated, reports are incomplete, and a victim cannot obtain a hearing. A lawsuit against an individual officer does not provide a durable remedy if doctrine asks the victim to identify a prior judicial opinion involving nearly identical facts. A damages system that exposes every reasonable mistake to personal ruin will also distort split-second decisions and make recruitment harder. The design must reach the government that authorized, trained, supervised, and insured the work while preserving consequences for deliberate abuse.

A national use-of-force floor

Congress should condition federal law-enforcement assistance on a clear floor for force policy, training, reporting, and review. The floor should apply directly to federal officers and to state, local, and tribal agencies accepting covered federal funds. States may adopt stronger rules.

Force should be authorized only to accomplish a lawful objective when no reasonably effective, safe, and feasible alternative appears available. Its level and duration must be proportionate to the threat, resistance, and objective. Deadly force should require an imminent danger of death or serious physical injury. Officers should give a warning when feasible, stop force when the need ends, obtain or render medical aid, and intervene when another officer is using unlawful force. De-escalation is required when objectively feasible and when it would not increase danger to an officer or another person.

This is a discipline standard, not a demand that officers wait to be injured. “Feasible” matters. The law must be judged from the circumstances a reasonable officer faced, including time, distance, weapons, bystanders, available cover, the severity of the suspected offense, active resistance, and the capacity of other responders. The standard becomes credible through scenario training that includes lawful force, successful de-escalation, failed de-escalation, medical response, crowd management, domestic violence, disability, behavioral crisis, and the duty to stop a colleague.

Chokeholds and carotid restraints should be treated as deadly force. No-knock entry should require a judge's specific finding of imminent danger that cannot be managed through a less dangerous method, and an on-scene supervisor should reassess the plan when circumstances change. Shooting at moving vehicles should be barred unless the vehicle or another weapon poses an imminent lethal threat that cannot be avoided. Policies should address foot pursuits, prone restraint, conducted- energy weapons, canines, crowd-control munitions, and force against a person whose danger is only to themselves.

Reporting that can answer a question

The federal government has operated a National Use-of-Force Data Collection since 2019, but agency participation remains voluntary. In its August 2025 release, the FBI reported that submitting agencies covered 78 percent of the law-enforcement population. Because that was below the 80-percent release threshold, it did not publish a national incident count. A democracy should not need to guess how often its agents use the most serious force.

Covered agencies should submit a zero report when no qualifying event occurred and an incident record when one did. Required fields should include the objective of the encounter, suspected offense, type and sequence of force, injuries, weapons, aid, warning, de-escalation, officer and civilian demographics, body-camera availability and activation, location, arrest disposition, and the status of administrative, criminal, and civil review. Public data should protect victim and witness identities while permitting analysis by agency, geography, event, outcome, and time.

The reporting obligation should extend beyond force. Stops, searches, arrests, deaths in custody, complaints, sustained findings, discipline, decertification, and departures during investigation are needed to distinguish a violent incident from an institutional pattern. Agencies must publish their policy, union contract, discipline matrix, body-camera rules, complaint path, reporting completeness, and correction log in accessible formats. Data should be versioned and machine- readable under PRIV-03, with privacy thresholds that do not expose a victim in a small community.

Federal money supplies leverage. A late or incomplete report should first trigger technical assistance and a cure period. Persistent refusal should redirect discretionary grants away from unreported equipment or operations and toward compliance, victim services, or a state reporting entity. Emergency communications and life-safety funds should not be abruptly removed from a community because its leadership failed. The consequence should reach the noncompliant function.

Independent review with a defined trigger

Police necessarily investigate most crime. They should not make the final institutional judgment about the gravest allegations against their own members. States receiving covered funds should maintain an independent serious-incident unit, assign a special prosecutor through a neutral rule, or use an equivalent structure separated from the involved chain of command.

Automatic referral should cover a death, life-threatening injury, officer-involved shooting, alleged sexual assault under color of law, intentional evidence destruction, and retaliation against a complainant or reporting officer. The independent body should control investigative scope, retain forensic experts, receive camera and communications data automatically, interview witnesses without the involved agency screening them, and publish a reasoned charging or non- charging decision subject to lawful redaction. It should not presume guilt. Independence protects the officer cleared by credible evidence as much as the civilian whose account would otherwise be judged by the accused institution.

Administrative review remains distinct. Conduct may violate policy or training without satisfying the criminal standard beyond a reasonable doubt. Civil liability has another standard. Outcomes must be described accurately: “not charged” is not “proved justified,” while a policy violation is not automatically a crime. Public reporting should show which question each process answered.

Replace qualified immunity with government responsibility

Section 1983 supplies a civil action against state and local officials who violate federal rights under color of law. Modern qualified-immunity doctrine generally shields an official unless the right was clearly established. In practice, litigation can turn on whether prior precedent stated the rule with enough factual specificity. That inquiry can defeat a claim even when the alleged conduct was unconstitutional. It also leaves the law undeveloped when courts resolve immunity without deciding whether a right was violated.

Congress should enact a cause of action for deprivation of constitutional rights by federal, state, local, or tribal law-enforcement officers acting under color of law and make the employing government the ordinary defendant. A claimant should prove the underlying violation and causation. The government should not defeat the claim merely because no earlier case involved materially identical facts. Ordinary defenses about reasonableness, lawful authority, factual mistake, causation, damages, and limitations remain available.

Damages should ordinarily be paid by the government or its risk pool when an officer acted within the scope of duty. This recognizes that policing is public work organized by public policy. It also lets insurers and governments price repeated misconduct, training failure, poor supervision, and bad deployment. Officers should face personal contribution only after a separate finding that the violation was intentional, knowing, or reckless, or that the officer acted for a personal purpose outside duty. Contribution should be proportionate and should not displace discipline, decertification, or criminal prosecution.

The statute should allow declaratory and injunctive relief, reasonable attorney fees, preservation orders, and equitable remedies for a continuing policy. It should set a fair federal limitations period, toll claims while a person is incapacitated or evidence is unlawfully concealed, and avoid damages caps that make severe injury administratively cheap. Frivolous claims can be dismissed under ordinary pleading, evidence, and sanction rules. “Frivolous” should not be a label that prevents discovery of records solely controlled by the government.

Body cameras as evidence infrastructure

Body-worn cameras can preserve evidence, resolve complaints, improve training, and show humane police work that would otherwise go unseen. They are not an automatic truth machine. A camera has a field of view, can miss events before activation, may not show what the officer perceived, and can turn private homes, medical crises, children, and victims into a government video archive.

Covered agencies should require activation for enforcement actions, searches, arrests, pursuits, transports, and use-of-force encounters, with stated exceptions for sensitive victim interviews or when recording would create a specific danger. Officers should state a reason on camera when they deactivate. Intentional nonactivation or deletion should have predictable evidentiary and disciplinary consequences; a technical failure or emergency mistake should not be treated as deliberate concealment.

Retention should follow purpose. Evidentiary and complaint-related footage remains available through the proceeding and appeal. Unflagged footage should expire on a short schedule. Facial recognition, generalized intelligence search, and unrelated immigration or protest surveillance should require separate legal authority. People recorded in an encounter should have a timely access path, subject to protection of victims, minors, medical information, and an active investigation. Public release after death or serious injury should follow a deadline, with a judge or independent reviewer able to approve a specific extension.

Capacity and specialization

“Fund” and “defund” are inputs, not operating models. Budgets should follow workload, service design, and measured outcomes. Some departments need more patrol officers, detectives, dispatchers, supervisors, evidence technicians, and training. Some have armed officers handling calls better served by civilian traffic responders, behavioral-health teams, victim advocates, or code and outreach staff. Many need both: enough police for violence and enough alternatives that police are not the only institution open at 2 a.m.

Federal grants should require a public workload and capability plan. The plan should map call type, priority, response time, staffing, overtime, case assignment, clearance, force, complaints, training, retention, and alternative-response availability. Communities may choose different models, but they should show who answers each call, what authority that responder has, how handoffs work, and what happens when a scene becomes dangerous.

Violence interruption, focused deterrence, environmental design, youth opportunity, domestic- violence intervention, reentry, and hospital-linked services can prevent harm before an arrest. They should be funded and evaluated without pretending every promising program replaces general policing. A program earns continued support through defined outcomes, comparison where feasible, community safety, participant rights, and an exit rule when it fails.

Fair accountability for officers

Accountability without due process is another form of arbitrary power. An officer accused of misconduct should receive the allegation, relevant evidence, representation, an opportunity to answer, a neutral and timely decision, written reasons, and an appeal. Investigators must disclose exculpatory material as well as inculpatory material. Interim reassignment can protect an investigation or the public without declaring guilt.

The process must still reach a conclusion. Contracts and statutes should not require destruction of sustained serious-misconduct records, impose deadlines too short to investigate concealed conduct, or permit resignation to erase a pending finding. States should license or certify officers and decertify for defined serious misconduct after due process. A hiring agency should see sustained findings, decertification, and a departure during a serious investigation. Minor and unfounded complaints should not become a permanent public accusation.

Leaders also owe officers a safe reporting path. The duty to intervene cannot work if the officer who stops misconduct becomes the easiest person to punish. Anti-retaliation rules, confidential reporting, independent investigation, and career protection should treat truthful intervention as professional courage.

The boundary with adjacent issues

JUS-01 owns bail, sentencing, rehabilitation, and algorithmic justice. This issue owns the authority, data, review, and remedy surrounding police conduct. CONST-04, CONST-05, and CONST-06 supply search, due-process, and criminal-procedure constraints; JUS-02 turns those rights into an operational policing contract. HEALTH-05 owns clinical crisis response and involuntary-care safeguards. JUS-02 requires police to remain available for violence, weapons, medical danger, and unsafe scenes while supporting a real clinical alternative.

CIVIC-04 addresses equal protection and repair across institutions. JUS-02 applies that demand to stops, searches, force, service, and remedy without treating every disparity as proof of an individual officer's intent. PRIV-03 governs public accountability data. This issue specifies the policing records that must exist, the privacy boundaries around them, and the enforcement hook that makes submission real.

The conclusion is deliberately not a midpoint. Police receive lawful power and the capacity to use it. Residents receive a rule that follows the force, evidence that can test the rule, an independent judgment in the gravest cases, and a civil remedy against responsible government. Neither side receives immunity from evidence.