Probation and Parole

Two decisions, two decision-makers

Probation is a sentence. In the federal system, 18 U.S.C. 3551(b) lists a term of probation alongside a fine and a term of imprisonment as one of the three sentences a court may impose on an individual, and state codes are built the same way. The judge decides at sentencing, and the person generally serves the term in the community rather than in custody. In many jurisdictions probation accompanies a suspended sentence, so a violation can trigger the confinement the judge originally held back.

Parole is a release decision, not a sentence. It happens after a person is already serving a prison term, and it is made not by the sentencing court but by a paroling authority: a parole board, a parole commission, or in some states a corrections official operating under statutory release rules. Parole does not shorten the sentence — it moves the remainder into the community under supervision, and the balance can be reimposed if supervision fails. Names vary by state, and several states abolished discretionary parole release while keeping a mandatory supervision period after prison, so 'does this state have parole?' rarely has a one-word answer.

Day to day the two statuses can look identical: similar conditions, similar reporting, sometimes the same agency and officer. The difference is jurisdictional, and it governs everything downstream — who set the conditions, who can modify them, and who hears a violation. This lesson is educational material about how these systems are structured. It is not legal advice, it describes U.S. jurisdictions only, and the governing rules are state-specific.

The federal exception: supervised release is not parole

The Sentencing Reform Act of 1984 ended federal parole prospectively. The effective-date provision enacted with it takes the sentencing chapter live on the first day of the first calendar month beginning 36 months after the October 12, 1984 enactment — November 1, 1987 — and applies it only to offenses committed after it takes effect. BJS states the consequence plainly: people sentenced to federal prison are no longer eligible for parole. The U.S. Parole Commission did not disappear, because people sentenced for pre-November-1987 conduct remain parole-eligible, as do D.C. Code offenders, military prisoners, and people transferred to the United States under prisoner-transfer treaties.

What took parole's place is supervised release under 18 U.S.C. 3583. Three features separate it from parole. The sentencing judge imposes it as part of the original sentence, and in some circumstances must; a board does not grant it later. It is served after the prison term is complete, so it adds supervision rather than substituting community time for confinement. And its maximum length is fixed by offense class — up to five years for a Class A or B felony, three for a Class C or D felony, one for a Class E felony or a non-petty misdemeanor. If the court finds a violation by a preponderance of the evidence, section 3583(e)(3) lets it revoke and require the person to serve part or all of the term in prison.

A measurement trap follows. The BJS Annual Parole Survey counts federal supervised release inside its federal parole figures, because it groups post-prison community supervision together, so national parole totals include a population that legally is not parole. That is a reporting convention, not a legal claim — but worth knowing before quoting a parole number.

Conditions, and the officer who both helps and enforces

Every supervision order is a list of conditions in two layers. Standard conditions apply to essentially everyone under that jurisdiction's supervision; special conditions are tailored to the individual case. Federal probation shows the structure cleanly. Under 18 U.S.C. 3563(a) the court must impose conditions including that the person commit no other federal, state, or local crime, not unlawfully possess a controlled substance, submit to drug testing, make restitution and pay assessments, and notify the court of material changes in economic circumstances affecting the ability to pay. Section 3563(b) supplies discretionary conditions the court may add when reasonably related to the sentencing factors: employment or training, treatment, reporting to a probation officer, residence and travel limits, community service, electronic monitoring.

The supervising officer holds two jobs at once. The federal judiciary describes post-conviction supervision officers as working both to protect the community and to help supervised people make long-term positive changes, using individualized assessment to decide how intensively each case is supervised. That dual role is real, not rhetorical: the same officer who arranges a treatment slot writes the violation report, and supervised people know it, which shapes what they disclose about a relapse, a lost job, or an unpaid fee.

Caseload is where the role gets squeezed. The American Probation and Parole Association published recommended adult ratios in 2006 — 20 cases per officer for intensive supervision, 50 for moderate-to-high risk, 200 for low risk — while insisting the numbers are a starting point rather than a standard, since caseload counts cases assigned while workload measures the time those cases actually consume. APPA also noted that average caseloads in many jurisdictions run much larger than experienced practitioners would recommend. These are professional recommendations, not legal requirements, and they are now two decades old. The volume is not in doubt: BJS estimated 1,626,300 entries onto probation and 1,607,500 exits from it during 2023 alone.

Violations, revocation, and the constitutional floor

Violations come in two kinds. A technical violation is a breach of a condition that is not itself a crime: a missed appointment, a positive drug test, leaving the district without permission, an unpaid supervision fee. A new offense violation is fresh criminal conduct, which can be handled both as a violation and as a separate prosecution — and the violation forum uses a lower standard of proof than the criminal case does.

Revocation is not a trial, but it is not unconstrained either. Morrissey v. Brewer, 408 U.S. 471, decided June 29, 1972, held that revoking parole requires two stages: a reasonably prompt preliminary inquiry before a neutral hearing officer near the place of the alleged violation or arrest, and a final revocation hearing. Chief Justice Burger's opinion set six minimum requirements: written notice of the claimed violations, disclosure of the evidence, an opportunity to be heard and to present witnesses and documents, confrontation and cross-examination of adverse witnesses unless the hearing officer specifically finds good cause otherwise, a neutral and detached hearing body whose members need not be judges or lawyers, and a written statement of the evidence relied on and the reasons for revoking. The Court was explicit that this is not to be equated with a criminal prosecution and should stay flexible enough to consider letters and affidavits a trial would exclude, and it expressly declined to decide the counsel question.

Gagnon v. Scarpelli, 411 U.S. 778, decided May 14, 1973, took up what Morrissey left open. Justice Powell's opinion extended the two-hearing requirement to probation revocation on the same terms, then made appointed counsel a case-by-case determination for the hearing body rather than a flat rule. Counsel should presumptively be provided where an indigent person makes a timely and colorable claim that he did not commit the violation, or where the violation is uncontested but substantial reasons in justification or mitigation make revocation inappropriate; refusals must be explained in the record. The reasoning turned on the differences from a trial: the state is represented not by a prosecutor but by a supervision officer, formal rules of evidence do not apply, and the hearing body's task is predictive and discretionary as well as factfinding.

Bearden v. Georgia, 461 U.S. 660, decided May 24, 1983, put a floor under money conditions. Bearden was placed on probation conditioned on paying a $500 fine and $250 in restitution, paid $200, was laid off, could not find other work despite trying, and was revoked to prison. Justice O'Connor's opinion held that a court cannot revoke for failure to pay without evidence and findings that the probationer was responsible for the failure, or that alternatives would be inadequate.

What the national numbers show

Scale first. At yearend 2023 BJS estimated 3,772,000 adults under community supervision — 3,103,400 on probation and 680,400 on parole — against 1,852,900 held in state or federal prisons and local jails. Roughly twice as many people were under supervision as were confined. BJS also reports that one in 70 adult U.S. residents was under community supervision during 2023, and that among agencies reporting comparably the total fell 23 percent over the decade to yearend 2023.

Outcomes next. Among 2023 probation exits with a known exit type, 58.5 percent were successful completions and 15.9 percent were exits to incarceration — totals that hide a large split. Misdemeanor probationers completed at 69.7 percent against 49.5 percent for felony probationers, and felony probationers left for incarceration at more than twice the misdemeanor rate (21.4 against 9.2 percent). Among parole exits with a known type, 63.9 percent completed and 28.6 percent returned to incarceration. In the reported parole detail, 73,680 exits were returns to incarceration and 47,660 of them — about 65 percent — were returns with revocation rather than with a new sentence.

Finally, the flow back into prison. Of 472,278 admissions of sentenced prisoners in 2023, BJS classified 111,385 (24 percent) as conditional supervision violations — returns from post-custody community supervision — and 350,628 (74 percent) as new court commitments. Read that 24 percent carefully: BJS defines new court commitments to include probation violators sentenced by a court, so people revoked from a straight probation sentence sit in the 74 percent, not the 24. The supervision-violation share is therefore larger than 24 percent, and BJS's national tables do not isolate the total. Reporting also varies — Florida does not report prison admissions for technical violations at all.

What the evidence supports, and where it does not

The dominant framework for organizing supervision is risk-need-responsivity, developed in the 1980s and first formalized in 1990 by Donald Andrews, James Bonta, and Robert Hoge. Risk: match supervision and service intensity to assessed likelihood of reoffending, concentrating resources on higher-risk cases. Need: target changeable factors linked to criminal conduct — substance use, procriminal attitudes, antisocial associates — rather than needs that are real but not criminogenic, such as self-esteem. Responsivity: use cognitive social learning methods, adapted to the individual's learning style, motivation, and circumstances. Most modern risk-assessment instruments are attempts to operationalize these three ideas.

Graduated sanctions are the other half of contemporary practice: instead of tolerating violations until one final breach triggers revocation to a long prison term, respond to every violation with a small, immediate, predictable consequence. The best-known version is the swift, certain, and fair model built around Hawaii's HOPE program, begun in 2004 under Judge Steven Alm, which paired a formal warning hearing with frequent random drug testing and brief jail stays for each violation. Early Hawaii evaluations reported large gains, and the model spread widely.

The replication record is where honesty matters. The National Institute of Justice funded a four-site randomized controlled trial — the HOPE Demonstration Field Experiment — in Arkansas, Massachusetts, Oregon, and Texas, enrolling 1,504 probationers randomly assigned to HOPE or probation as usual. The 2018 final report found no reduction in recidivism: 40 percent of HOPE participants versus 44 percent of the comparison group were rearrested, 25 versus 22 percent were revoked, and 28 versus 26 percent were reconvicted. HOPE participants were more likely to go to jail (82 versus 56 percent), served more total jail days (47 versus 33), and cost more. The evaluators did not conclude that swift-certain-fair supervision should be abandoned; they noted that two sites already had low revocation rates leaving little room to improve, and that the Hawaii model as its designer describes it layered swift sanctions on top of an RNR and motivational-interviewing approach the trial did not test. A well-replicated null result should temper large claims rather than settle the matter.

Financial conditions deserve their own line, because failing them is a documented reason supervision fails: in the HOPE trial, 18 percent of the probation-as-usual group and 11 percent of the HOPE group were violated for not paying fees and fines. The U.S. Commission on Civil Rights reported in 2017 that nonpayment of legal financial obligations can trigger driver's license suspension in at least eight of the fifteen states with the largest prison populations, and that some states incarcerate people or extend probation or parole terms for nonpayment. Bearden limits only the sharpest version of this: it requires an inquiry into responsibility for nonpayment before revocation, and it does not erase the underlying obligation.