Criminal Justice Reform

Cash Bail — State Authority, Federal Guardrails

Cash bail remains a state authority — this is not a federal ban or mandate. The federal role is narrower: new guardrails specifically targeting 501(c)(3) charitable bail funds to prevent systemic abuse of the pretrial release mechanism. States decide their own bail frameworks. The federal government ensures that charitable bail organizations operate within their intended purpose and are not used to circumvent the system wholesale.

Policing Accountability

Federally-funded mandatory body cameras for all law enforcement officers, with public footage-release timelines written into the funding requirements — not left to departmental discretion. A national police misconduct and decertification database tracks officers across state lines, closing the loophole that allows decertified officers to move to a new jurisdiction and be rehired. A bad officer who loses their badge in one state cannot simply pick it back up in another.

Sentencing — Truth in Sentencing with Rehabilitation Inside

Truth-in-sentencing: time served matches the sentence handed down. Less parole-board discretion, more predictability — for victims, for defendants, and for the system. Combined with mandatory education and job-training programming delivered during incarceration, in secure settings — not as a diversion or an alternative to incarceration, but as part of it. The goal is not to soften sentences. It is to ensure that the people who serve them are less likely to return.

There's a real internal tension worth acknowledging rather than ignoring: truth-in-sentencing removes the parole-board discretion that traditionally gives incarcerated people an incentive to participate in rehabilitation programming — if time served is fixed regardless of behavior, what's the incentive to engage? The answer is that program participation itself becomes a condition of eligibility for the specific privileges truth-in-sentencing doesn't touch — visitation, work-release eligibility, facility placement — rather than the release date itself, which stays fixed either way.

Federal Law Enforcement Accreditation — Built on What Already Works

A real, credible national accreditation standard for police departments already exists — the Commission on Accreditation for Law Enforcement Agencies (CALEA), founded in 1979 by the International Association of Chiefs of Police, the National Sheriffs' Association, and other major law enforcement executive associations. CALEA's 185-standard framework covers written use-of-force policy, data-driven management, training, and operational accountability. It's a genuine, respected professional standard. It's also voluntary, and adoption reflects that: roughly 800 of the nation's approximately 18,000 law enforcement agencies are actually accredited — under 5%.

This platform doesn't reinvent that standard. It adopts CALEA's existing framework as the federal baseline, using CALEA's own established assessment and audit infrastructure rather than building a new federal bureaucracy to duplicate work that's already being done credibly. What changes is that it stops being optional: accreditation to this baseline becomes a condition of federal law enforcement funding and support — COPS Office hiring and equipment grants, DOJ discretionary grants including Byrne JAG funding, and participation in the DoD 1033 surplus equipment transfer program. No accreditation, none of that federal support. This is the same fiscal-accountability logic already applied elsewhere in this platform to state grant funding (see Public Trust): federal dollars come with a real, verifiable standard attached, not a formality.

The honest complication, and the reason this can't just be a blunt mandate: accreditation costs money — application and annual fees, assessor costs, dedicated compliance staff time — and cost is the most commonly cited reason departments don't pursue CALEA accreditation voluntarily today. A hard, immediate, unfunded mandate would hit small and rural departments hardest — the ones most dependent on COPS grants and 1033 equipment transfers in the first place — and risk the opposite of the intended effect: federal law enforcement support concentrating in large, already well-resourced departments that can absorb compliance costs, while small agencies lose federal funding entirely rather than gaining better standards. The fix: a funded phase-in period, with federal cost-sharing specifically for departments below a defined size or budget threshold, so the standard doesn't become a de facto tax on small-town policing. This isn't a footnote — it's the difference between a policy that raises standards nationally and one that just defunds rural law enforcement while calling it accountability.

Accreditation status for every federal-funding-eligible department is published on the same real-time Public Transparency Dashboard already proposed elsewhere in this platform — searchable by any citizen, the same way federal spending and judicial accountability data are, rather than requiring a records request to find out whether a local department meets a real professional standard.

Building Federal Infrastructure, Not Just Requiring Private Accreditation. This requires an act of Congress, not executive action — real federal appropriations to build and operate the accrediting body and its supporting database, not a directive issued unilaterally. That's the honest scope of the ask. In exchange, a federally-funded, federally-run accrediting body can do something CALEA's private, fee-funded model structurally can't: operate without charging individual departments per-agency accreditation fees, which is the actual mechanism driving the rural-department cost burden described above. A federal database also lets the standard go beyond CALEA's baseline where this platform has a specific reason to require more.

One standard this platform sets beyond CALEA's existing framework: mandatory initial and recurring mental health evaluations for sworn officers, administered by licensed professionals and structured to comply with the ADA — job-relevant fitness assessment, not general psychiatric labeling, with confidential handling of results consistent with existing professional guidelines already used in pre-employment screening today.

The initial evaluation requirement isn't new in concept — most states already require pre-employment psychological screening for peace officers through their POST commissions. What's new is making it a uniform national baseline rather than a patchwork that varies by state. The recurring evaluation requirement has a real federal precedent behind it: VA Directive 0730 already requires periodic psychological evaluations for VA Police Officers — the federal government's own police force. This platform extends a model the federal government already uses on itself to every department seeking federal accreditation, rather than asking local departments to do something federal law enforcement doesn't already do.

[PENDING: The exact recurrence interval isn't set here — VA's own directive establishes one for its officers, and the federal accrediting body should set the interval for this program specifically, informed by that existing federal model rather than an arbitrary number chosen without it.]

The confidentiality and non-punitive framing matters as more than a courtesy: occupational mental health research consistently finds that evaluation programs perceived as purely punitive drive underreporting, undermining the program's actual purpose. Evaluations flagging a concern route to a support and treatment pathway by default, not automatic removal from duty — consistent with how ADA-compliant fitness-for-duty processes already have to work today. To be precise about scope: this is about consequences for the individual officer regarding evaluation results specifically. Department-level consequences for failing accreditation generally are a separate, harder mechanism — loss of federal funding eligibility, described above, with no equivalent softening.

Additional Accreditation Standards. Beyond mental health evaluation, the federal baseline requires: recurring in-service training, not just initial academy certification — an officer's qualification doesn't freeze at hiring; a specific, standalone de-escalation curriculum, distinct from general use-of-force training; and firearms proficiency requalification on a recurring schedule, not a one-time academy qualification treated as permanent.

On background investigation specifically: federal law already prohibits anyone convicted of a misdemeanor domestic violence offense from possessing a firearm for life, under the Lautenberg Amendment (18 U.S.C. § 922(g)(9)) — and Congress specifically removed the law enforcement exemption that existed before Lautenberg passed. There is no carve-out for police officers or their service weapons. On paper, this problem is already solved. In practice, it isn't, for two reasons: inconsistent background-check rigor at hiring across roughly 18,000 separate departments, and no reliable mechanism to catch a disqualifying conviction that happens during employment rather than before it — particularly given the well-documented pattern of officer-involved domestic violence being under-charged or handled informally compared to identical conduct by a civilian, which can prevent the conviction that would trigger Lautenberg from ever happening in the first place.

Federal accreditation requires a standardized background investigation protocol at hiring, and — this is the actual fix, not the restated existing law — continuous cross-checking of serving officers against disqualifying convictions and protective orders for the duration of their employment, not a background check performed once and never revisited. A Lautenberg-disqualifying conviction, whenever it occurs, triggers automatic entry into the same national police misconduct and decertification database this platform already establishes elsewhere in Criminal Justice Reform — closing the loop between a federal firearms disqualification and actual removal from armed duty, rather than relying on the officer's own department to notice, report, and act on its own.

Overcrowding Is a Constitutional Problem, Not Just a Budget Problem

The US incarcerated population sits at roughly 1.9 million and is growing again after a pandemic-era decline — state prisons alone added more than 50,000 people over 2022-2023, undoing about a quarter of that reduction. The total system costs at least $182 billion a year. Crowding isn't uniform: Iowa runs at 119% of rated capacity, the worst in the country, while other states sit well under their limits.

This isn't just a resourcing problem — the Supreme Court has already ruled it can be an independent constitutional violation. In Brown v. Plata (2011), the Court upheld a lower-court order forcing California to reduce its prison population from roughly 200% to 137.5% of design capacity — a release of 38,000 to 46,000 people — because overcrowding itself was found to be the primary cause of medical and mental health care so inadequate it violated the Eighth Amendment. That's the legal floor this plank works from: overcrowding isn't just expensive, it's a constitutional problem in its own right once it degrades care below that standard.

The mechanism here isn't a new bureaucracy — it's two pieces already elsewhere in this platform, applied to population reduction directly. Cash bail reform and Drug Policy's civil-commitment diversion both reduce population at the actual source, since most jail populations are unconvicted pretrial detainees, not sentenced inmates.

On top of that, and deliberately designed not to undermine truth-in-sentencing elsewhere in this platform: judicial — not parole-board — second-look review for elderly and medically fragile inmates, available only after a long fixed minimum of 15-20 years already served. Eligibility is triggered by objective factors alone — documented age and medical status — not behavior credit or discretionary good-time, so this doesn't reopen the early-release-for-good-behavior door truth-in-sentencing closes elsewhere. It targets the specific population that is simultaneously lowest-risk to release and most expensive to continue incarcerating.

[NOTE: The research on forced population reductions and recidivism is genuinely mixed. UC Irvine's study of California's post-Plata realignment found no increase in major crimes, but also found the population actually diverted was higher-risk than intended, because the vague "county supervision" category used at the time let higher-risk cases through. This plank's objective age/medical criteria are specifically designed to avoid that failure mode by removing discretion from the eligibility call — but that design choice hasn't been independently verified as sufficient on its own.]

Expanded Work Release and Job Training

Federal Prison Industries — UNICOR — already exists, is already proven effective, and is already starved of capacity. The DOJ's own Post-Release Employment Project study found UNICOR participants were 24% less likely to reoffend within 12 years of release and 14% more likely to be gainfully employed afterward. It's also self-funding — UNICOR sells to federal agencies and receives no taxpayer appropriation for its operations. And yet only 8% of work-eligible federal inmates actually hold a UNICOR position; roughly 25,000 sit on a waiting list for a slot that doesn't exist. This platform's position: fund UNICOR's expansion specifically to close that waiting list, prioritized for inmates with less than 10 years remaining on their sentence — the population for whom job-ready skills at release matter most directly, and where the reentry-outcome evidence is strongest.

Mandatory Cross-Cultural Group Therapy and Harm Reduction Classes

Cross-cultural group therapy and harm reduction education for every inmate, not just those who opt in. One honest design tension worth naming rather than glossing over: mandating group therapy participation for an entire population — including people who don't want to be there — runs against standard clinical practice, where group therapy effectiveness depends heavily on genuine participant engagement, and mandatory attendance can undermine the experience for participants who do want to engage. The fix isn't abandoning the mandate, it's distinguishing what's mandatory from what's clinically dependent on willingness: attendance and participation in the curriculum is mandatory; therapeutic depth of engagement isn't something a mandate can force, and the program has to be structured by corrections mental health professionals with that distinction in mind, not designed as if compulsory attendance alone guarantees compulsory therapeutic benefit.

Harm reduction classes — overdose recognition, naloxone use, safe practices — apply the same health-first philosophy already established in this platform's Drug Policy section to the incarcerated population directly, not just to civilians. This matters concretely: people leaving incarceration face a well-documented, sharply elevated overdose risk in the weeks immediately after release, driven in part by reduced tolerance after a period of enforced abstinence. Harm reduction education during the sentence, not first encountered at release, addresses that risk directly.

Mandatory Drug Screening During Sentence

The Bureau of Prisons already runs a drug testing program — but it's random-sample and risk-targeted (prior drug history, community program participation, specific suspicion), not universal testing of every inmate on a regular schedule. Full mandatory testing currently applies only at Community Treatment Centers, the halfway-house stage near release. This platform expands that to comprehensive, regular screening for the full sentence, not just the final stage — both as a security and facility-safety measure and as an early-detection tool feeding into the treatment and harm-reduction programming above, rather than relying on random sampling to catch what a comprehensive program would catch directly.

Housing by Rehabilitation Potential, Not One-Size-Fits-All

Facility housing assignment incorporates rehabilitation and release potential as a distinct classification factor, not just the security-risk score BOP currently uses alone. This deliberately sets aside a more aggressive alternative that was considered: consolidating the longest-sentence population into a single national facility. That approach was set aside for two real reasons — it would be a genuinely unprecedented construction undertaking (no current US facility approaches the scale involved, roughly 27,000 combined long-sentence inmates), and it risks cutting inmates off from family contact entirely, which research consistently links to worse institutional and mental health outcomes. This works within and across existing facilities instead — grouping inmates with realistic release and rehabilitation prospects together, where programming, work release, and job training resources concentrate, and housing the population that will not be released separately, where different programming priorities (the mental health, harm reduction, and long-term stability programming above) apply instead. This achieves the same basic goal as consolidation — matching resources and programming to the population that will actually use them for reentry — without the unprecedented scale and family-contact costs a single mega-facility would carry.

Eliminating the For-Profit Prison System

No one in the United States profits from another person's incarceration. This is a distinct rule from banning private ownership outright — a private entity may own or operate a facility, but the arrangement has to be structured so no one extracts wealth from doing it. Concretely: any private operator is organized as a nonprofit, not a for-profit corporation with shareholders or equity holders who receive distributions; contracts are cost-reimbursement only, not a per-diem or fixed rate that can generate a margin above actual cost; and compensation for anyone involved is subject to real oversight, not just a change of corporate label — a nonprofit structure alone doesn't guarantee no one profits from it if executive compensation becomes a disguised version of the same thing.

This has to be actual legislation, not another executive order: federal private prison policy has reversed three times in under a decade — phased out in 2016, reinstated in 2017, banned again in 2021, reinstated again in 2025 — and an executive order lasts exactly as long as the administration that signed it. The original 2021 ban was grounded in real findings, not just preference: DOJ's own Inspector General investigations found private facilities less safe and more likely to rely on harmful practices including solitary confinement.

The rule has to close a specific, already-documented loophole or it repeats a failure that already happened once: even under the 2021 ban, private prison operators kept significant federal business by contracting with county governments, who then contracted with the U.S. Marshals Service on the companies' behalf — USMS kept roughly 20,000 of its 63,000 detainees in private facilities this way, and received at least seven formal exceptions to the ban for exactly this arrangement. A for-profit company routing through a nonprofit shell or a county intermediary to keep extracting value doesn't comply just because the paperwork changed — the rule prohibits the practical effect, wealth extracted from someone's incarceration, regardless of which government layer or corporate structure sits on top of it.

This principle extends past facility ownership to every service a captive prison population has no alternative but to use: telecommunications, commissary, and contracted security services. The mechanism is 'street pricing' — an already real, currently-debated policy concept (Congressional inquiry into stadium and airport concessions used this exact term in 2025) applied here directly: a phone call, a commissary item, or a contracted service inside a correctional facility cannot cost more than the same call, item, or service costs on the open market outside it. Prison telecom and commissary companies currently charge captive populations dramatically more than street price precisely because there's no alternative vendor to switch to — the same captive-market dynamic driving the stadium and airport concession debate, except the audience here didn't choose to be there and can't leave to get a better price down the block. If it wouldn't cost more outside, it doesn't cost more inside.

Enforcement is a private right of action, held by the incarcerated individual or a family member who paid the overcharged price, and it's built correctly from the start rather than left to be fixed after the fact. Washington State's own pay transparency law is the direct lesson here: its original 2023 enforcement mechanism combined a private right of action with an automatic flat statutory damage, no chance for the vendor to correct the violation first — and the result, well-documented by 2025, was a wave of lawsuits driven by opportunistic plaintiffs seeking the automatic payout rather than genuine enforcement of the underlying protection. Washington fixed this in 2025 with a 5-business-day cure period and discretionary damages ($100–$5,000 per violation, weighed by vendor size, willfulness, and frequency) replacing the flat fine. This platform's street pricing enforcement adopts that already-tested structure directly: written notice to the vendor, a 5-business-day window to correct the pricing before any liability attaches, and discretionary statutory damages within the same $100–$5,000 range if uncorrected — calibrated deliberately above the size of a typical individual overcharge (a phone call or commissary item), since the deterrent value of enforcement depends on the penalty exceeding the harm, not just matching it.

This also has to explicitly include immigration detention, which the 2021 ban never touched at all — and it's the larger problem now, not the smaller one. ICE contracts account for 43% of GEO Group's revenue and 30% of CoreCivic's, compared to 3% and 2% respectively from the Bureau of Prisons. A policy that eliminates for-profit federal criminal incarceration while leaving for-profit immigration detention fully intact isn't closing the for-profit prison system — it's closing the smaller half of it.

Qualified Immunity Ends Where Deliberate Indifference Begins

Qualified immunity remains available for genuine good-faith mistakes — that protection exists for a real reason, established in Harlow v. Fitzgerald (1982): officials shouldn't face the burden of litigation, not just liability, for reasonable judgment calls made without the benefit of hindsight. This platform doesn't eliminate that protection wholesale.

It ends where deliberate indifference begins. A law enforcement officer who acts with deliberate indifference to a clearly established constitutional right — the same standard already used and court-tested in other constitutional tort contexts (Farmer v. Brennan, 1994) — has no qualified immunity defense. This is a narrower, more precise standard than 'knowing' violation: it doesn't require proving what was subjectively in an officer's head at the moment of the incident, which is exactly the kind of expensive, case-defeating inquiry the Supreme Court eliminated qualified immunity's predecessor doctrine to avoid. It asks instead whether the officer was aware of a substantial risk that their conduct violated a clearly established right and disregarded it — provable through objective evidence, including training records, not just testimony about intent.

That evidentiary link isn't incidental — it's the mechanism that makes this provable rather than aspirational. This platform's federal law enforcement accreditation program, established elsewhere in Criminal Justice Reform, already requires documented, recurring training on constitutional standards. An officer specifically and recently trained that certain conduct violates the Constitution, who engages in that exact conduct anyway, has left a paper trail establishing deliberate indifference — not a subjective guess about what they were thinking.

Liability is personal, not institutional, and that distinction is built to actually mean something rather than function as a label. Qualified immunity has always been an individual officer's defense, not the department's — a separate doctrine (Monell v. Department of Social Services, 1978) governs when a municipality itself can be sued, requiring proof of an official policy or custom, and that stays untouched here. What changes is what happens once a court — not the officer's own employer — finds deliberate indifference. Colorado's 2020 reform (SB 20-217) is the leading state precedent, and it's genuinely instructive both for what it got right and where it fell short. It created a real state cause of action free of qualified immunity — a real advance. But the employer decides whether the officer acted in bad faith, a direct conflict of interest, since the same department has every incentive to find good faith and fully indemnify rather than expose its own officer and its own training record to personal liability. And even when bad faith is found, personal liability caps at 5% of the judgment or $25,000, whichever is less — described even by police-training commentators as 'GoFundMe money,' not a real deterrent.

This platform closes both gaps. The deliberate indifference finding is judicial, made by the same court or jury deciding the underlying constitutional claim — not the officer's own department. And once that finding is made, there is no cap and no indemnification: the department cannot pay the judgment on the officer's behalf, and neither can private insurance. This isn't a novel restriction — it's the same principle already applied to punitive damages in numerous states, including California, where the Insurance Code explicitly bars insurers from covering losses caused by an insured's willful act, precisely because allowing someone to insure against their own intentional wrongdoing defeats the purpose of holding them personally accountable for it. Colorado's own experience proves this gap is real, not theoretical: within months of SB 217's passage, the Colorado Fraternal Order of Police's legal defense fund began covering officers' personal exposure under the law in full, with no deductible — a private insurance market emerging specifically to absorb the liability the reform was designed to create. This platform's version doesn't leave that door open.

Be direct about the practical limit here rather than overstate the fix: an individual officer's personal assets may still fall short of a large judgment, and no legal rule guarantees full collection in every case. That's true of personal liability generally, not a defect specific to this policy. The deterrent value comes from removing the guaranteed safety net — department indemnification and private insurance both — not from guaranteeing every plaintiff collects in full.

This requires federal legislation, since qualified immunity for federal constitutional claims is federal judicial doctrine that no state can override directly — the same reason Colorado's model, however real, only reaches state-law claims. Real bills already exist as a starting point: the Ending Qualified Immunity Act (H.R. 3602) and the Qualified Immunity Abolition Act of 2026 (S. 3625) both aim to eliminate the doctrine outright at the federal level. This platform's position is narrower and, it argues, more durable — preserving qualified immunity for genuine good-faith conduct while ending it precisely where deliberate indifference to a clearly established right can be proven, which is both a more targeted reform and a harder position for defenders of unconditional qualified immunity to argue against in good faith.

No More Policing for Profit — The FAIR Act, With a Victim-First Proceeds Rule

Civil forfeiture allows the government to permanently take your property — cash, car, home — without ever charging you with a crime, let alone convicting you of one. In 43 states, the law enforcement agency that seizes the property keeps anywhere from half to all of the proceeds, creating a direct financial incentive to police for profit rather than for public safety. Two separate Institute for Justice studies found no link between civil forfeiture and crime-fighting effectiveness — but did find that forfeiture activity ramps up specifically during periods of fiscal stress, when agencies need the revenue. That's the system working exactly as its incentives were designed.

The worst part isn't even the state-level problem — it's the federal escape hatch. Even the 15 states that have required a criminal conviction before forfeiture, and the 3 states (New Mexico, Maine, Montana) that have abolished civil forfeiture outright, can be bypassed entirely: local law enforcement simply routes seizures through a federal agency under the 'equitable sharing' program, the federal government processes the forfeiture under looser federal law, and the local agency collects up to 80% of the proceeds — bypassing the very state laws their own citizens voted for. Eric Holder restricted equitable sharing in 2015. Jeff Sessions reversed it in 2017. It's been running at full capacity ever since, a standing invitation to circumvent any state reform that inconveniences law enforcement financially.

This platform supports the FAIR Act (S. 263 / H.R. 7638, 119th Congress) — the bipartisan legislation sponsored by Rand Paul and Cory Booker, with ideologically diverse co-sponsors including Mike Lee and Angus King, that cleared the House Judiciary Committee 26-0 in a prior version. The FAIR Act requires a criminal conviction before any forfeiture can proceed, eliminates the equitable sharing loophole so federal law can't be used to circumvent stricter state standards, ends nonjudicial/administrative forfeitures that currently let agencies finalize seizures entirely internally without court involvement, and shifts the burden of proof to the government rather than requiring the property owner to prove their own innocence.

This platform adds one provision the FAIR Act doesn't include: a specified order of priority for forfeiture proceeds. First, proceeds go toward making victims of the underlying crime whole — restitution before anything else. What remains after victims are made whole goes to the general fund, not the seizing agency's own budget. Removing law enforcement's direct financial stake in forfeiture outcomes is the structural fix the FAIR Act's conviction requirement alone doesn't fully close — an agency that convicts someone and then receives the proceeds of that conviction still has a financial incentive to pursue forfeiture cases over other priorities. The profit motive has to be removed from the equation entirely, not just regulated.