Response cost · Atlas

1. Introduction

Version followed: the applied-behaviour-analytic procedure as defined in Cooper, Heron & Heward's Applied Behavior Analysis (2nd ed., 2007; the chapter on punishment by contingent removal of a stimulus), with the safeguards of the New Mexico DDSD Response Cost Guidelines (reviewed 2024) and the BACB Ethics Code for Behavior Analysts (effective 1 January 2022). Home use follows the token-system steps of Barkley's Defiant Children (3rd ed., 2013). Classroom variants follow the original JABA studies named below. [source: Cooper et al. 2007 via NM DDSD 2024; BACB 2022]

What it is. Response cost is not a stand-alone therapy, and it has no founder or manual. It is one procedure: the contingent loss of a set amount of a reinforcer the person already holds (points, tokens, minutes of free time, money, part of a privilege) after a defined target behaviour. It counts as response cost only if the behaviour then decreases. It is a form of negative punishment and is distinct from time-out: the person keeps access to the situation but loses some of what they hold. [source: Rapport & Begolli 2005; Cheney, Blum & Walker 2005; NM DDSD 2024]

How change happens. Losing a conditioned reinforcer contingent on a response reduces how often that response occurs. Weiner first showed this experimentally with humans losing points (1962). Kazdin reviewed applied uses in 1972. Clinically, it works only when the thing lost already has value, which in practice means a reinforcement system must be running. Its job is to suppress a problem behaviour while reinforcement builds the alternative behaviour. [source: Soares et al. 2017; Kazdin 1972 (cited) [uncertain]]

Terminology caution. In basic research, "response cost" is also used for increasing effort or ratio requirements. This guide covers only the clinical meaning, the loss of held reinforcers (token loss, fines). Soares et al. (2017) warn that the three meanings are not functionally equivalent. [source: Soares et al. 2017]

2. Framework

There is no manualised phase structure. The phases below are assembled from the usage guidelines in Cooper et al. (as summarised in the NM DDSD guidelines and course materials), the BACB code and Barkley's sequencing. [synthesis of named sources]

Phase 0: Decide whether response cost is justified.

Goal: confirm that a punishment-based procedure is warranted at all.
Tasks: complete a functional assessment (topography, antecedents, precursors, severity, likely function). Collect baseline data on frequency, severity and duration. Document that less intrusive, reinforcement-based procedures have been tried and the data show they failed. Under BACB 2.15, the only exception is when the team judges that the risk of harm from the behaviour outweighs the risk of the procedure. [source: NM DDSD 2024; BACB 2022, 2.14–2.15]
Criterion to proceed: baseline data exist, a reinforcement-based plan is in place or failed on data, and any required review (e.g., a human rights committee in disability services) has approved. [source: NM DDSD 2024; BACB 2.15]

Phase 1: Design.

Define the target behaviour(s) and what counts as one "incident".
Set the fine for each behaviour.
Choose the method: direct fine within an earning system, or bonus response cost from a reserve.
Set how immediately fines are applied.
Ensure the person cannot drop below zero.
Define the backup reinforcers and the exchange schedule.
Write the plan down.

[source: Cooper et al. guideline list as given in course flashcards; NM DDSD 2024; BACB 2.14]

Phase 2: Introduce. Explain the rules and obtain consent and assent (BACB 2.11; NM DDSD step 5). In parent training, the earning system runs first. Barkley's sequence introduces the poker chip/point system in Step 4, and adds disciplinary consequences in Step 5, a week later. [source: Barkley 2013 contents; CEBC listing] Whether Barkley tells parents to begin chip fines at Step 4 or Step 5 is [uncertain]. [uncertain]

Phase 3: Implement and monitor. Apply fines consistently and record every fine, the earnings and the target behaviour. Graph the data and review them. [source: BACB 2.17–2.18; Cooper et al. "keep records"]

Phase 4: Evaluate.

If the target behaviour is not clearly decreasing, it is not functioning as punishment: modify or stop. Continuing an ineffective potentially aversive programme may amount to abuse in disability services. [source: NM DDSD 2024; Cheney et al. 2005; BACB 2.15]
Disability services review at least quarterly. [source: NM DDSD 2024]

Phase 5: Fade.

Plan the exit from the start, because response cost programmes can persist for years. [source: NM DDSD 2024]
Reduce how often fines are used, keep reinforcement for the alternative behaviour, and keep addressing the underlying need. [synthesis]

Session count and format.

Classroom and residential: continuous, all day, with daily or weekly exchange.
Outpatient parent training (Barkley): 10 weekly sessions of 1 hour individually or 2 hours in groups. Token earning and fines cover about two of those sessions, and fines are then used at home between sessions.
Individual child CBT: fines can run within the session (see 4.9).

[source: CEBC listing of Barkley; Rapport et al. 1982 summary]

3. Guiding principles

Defined by effect. It is response cost only if the behaviour decreases. In practice: graph from day one, and if the data are flat after a fair trial, stop. [source: Rapport & Begolli 2005; NM DDSD 2024]
Reinforcement first and alongside. Prioritise positive reinforcement, and use response cost only after less intrusive means have failed. In practice: there must be more ways to earn than to lose, and the alternative behaviour is always being reinforced. [source: BACB 2.14–2.15; Cheney et al. 2005]
Address the need, not only the behaviour. In practice: keep working on the function the behaviour serves and on quality of life. Track preferred activities the person gets without earning or risking them. [source: NM DDSD 2024, citing Lovett 1996]
Define the behaviour and the fine in advance. In practice: a written list of behaviour → cost, so no fine is invented on the spot. [source: Cooper et al. via flashcards]
Decide how immediately fines are applied. In practice: fine as soon after the behaviour as feasible, or at a fixed review point. Pick one and keep to it. [source: Cooper et al. via flashcards]
Ensure a reinforcer reserve; no unrecoverable debt. In practice: the balance can never go below zero, and today's behaviour never wipes out last week's earnings. [source: Cooper et al. via flashcards; NM DDSD 2024]
Set magnitude adequately; do not ratchet up. In a residential trial, 30-token fines suppressed antisocial behaviour more than 5-token fines for 5 of 6 adolescents (Burchard & Barrera 1972). A small fine may be ineffective after a large one, and steep increases can damage the token economy. NM DDSD warns that raising fines to make them more powerful usually fails. In practice: pick a meaningful but recoverable fine at the start rather than escalating it later. [source: Burchard & Barrera 1972 summary; U Idaho notes; NM DDSD 2024]
Avoid overuse. In practice: if fines outnumber earnings for several days, the plan is wrong, not the person. [source: Cooper et al. via flashcards; U Idaho notes; NM DDSD 2024]
Expect unplanned outcomes. In practice: watch for aggression, avoidance of the person delivering fines, and collateral drops in desired behaviour. [source: Cooper et al. via flashcards]
No escape. In practice: the person must not be able to avoid the fine by leaving or refusing, so plan a backup consequence for refusing to pay. [source: U Idaho notes]
Never fine basic rights or needs. NM DDSD notes that Cooper et al. caution that removing structured or free time may be inappropriate for people with developmental disabilities. In practice: fine only true extras, never food, family contact, safety, therapy or rights. [source: NM DDSD 2024] [synthesis for the list of examples]
Consent, assent, review. In practice: document that the person understands the programme, and obtain consent and any required committee approval before starting or changing it. [source: BACB 2.11, 2.15; NM DDSD 2024]
Neutral, brief delivery. One listed drawback of response cost is that it calls attention to the punished behaviour. In practice: name the behaviour and the cost in one sentence, without lecturing or emotion. [source: Cooper et al. via flashcards (drawback); delivery advice is synthesis]
Would you accept it yourself? NM DDSD advises never applying to a person with a disability an intervention you would not accept for yourself. In practice: use this as a quick test of fairness. [source: NM DDSD 2024]

4. Intervention inventory

Cooper et al. list four response cost methods: direct fines, bonus response cost, combining with positive reinforcement, and combining with group consequences. The other variants below come from the applied literature and named programmes. None of them is a separate therapy. [source: Cooper et al. via course flashcards; Cheney et al. 2005; NM DDSD 2024]

4.1 Direct fines within a token or point economy (token loss)

Purpose / when: this is the core method. Use it to reduce a specific, well-defined behaviour in someone already earning tokens or points that they value. Settings include a classroom, residential unit or home. [source: Rapport & Begolli 2005; Cheney et al. 2005]
Steps:
1. Confirm the earning system has run long enough that tokens are clearly valued: the person exchanges them and asks about them. [synthesis]
2. List the target behaviours, each with an observable definition, and the fine for each. Post the list. [source: Cooper et al. guideline "define behaviour and fine"]
3. Define an "incident". NM DDSD gives an example: a cluster of threats separated from the next cluster by about 10–15 calm minutes counts as one incident with one fine, not 25. [source: NM DDSD 2024]
4. Set fines that are meaningful but recoverable within the same day or exchange period. [synthesis from Burchard & Barrera 1972 and NM DDSD]
5. Deliver the fine promptly: name the behaviour, state the cost, remove the token or mark the deduction. Do not argue.
6. Continue to deliver earned tokens for appropriate behaviour, including later in the same hour.
7. Record each fine (time, behaviour, amount) and each earning. Graph daily. [source: BACB 2.17]
Example language (illustrative): "That was hitting. That's two points." Later: "You've been working calmly for ten minutes. That's a point earned."
Watch for: protest, refusing to hand over tokens, and a balance near zero. A precedent is Achievement Place, the teaching-family home for adolescent boys, which used point fines inside a token economy (Phillips 1968). The specific behaviours fined there were not checked. [uncertain]
Working if: the target rate falls within days to a couple of weeks against baseline, earnings stay above fines, and no new problem behaviour appears. [synthesis; moderate-to-rapid decrease listed as an advantage in Cooper et al.]

4.2 Bonus response cost (reserve-based)

Purpose / when: the person is given a non-contingent reserve ("bonus") at the start of a period, and set amounts are removed for each instance of the target behaviour. Use it when there is no earning system, when earnings are too low to fine from, or for a short period such as one lesson or one outing. [source: Cooper et al. definition via flashcards]
Steps:
1. Choose the period (e.g., a class period, morning, car trip) and the reserve size. The reserve must cover the realistic number of incidents with something left over. [synthesis]
2. Make the reserve visible: tokens in a cup, tally marks on a card, minutes on a board.
3. Explain: "These are yours. You keep what's left at the end."
4. Remove a set amount per defined incident, using the delivery in 4.1.
5. At period end, exchange the remainder for the backup reinforcer immediately.
6. Restart the reserve fresh next period. Debt never carries forward. [source: NM DDSD 2024 on unrecoverable debt]
Example language (illustrative): "You start the bus ride with ten stars. Each time you unbuckle, one star goes. Whatever's left buys tablet time when we get home."
Watch for: the reserve running out early, after which there is nothing left to lose. If it runs out, shorten the period or reduce the fine. [synthesis] Also watch for the person treating the reserve as "free" if exchange is delayed too long. [synthesis]
Working if: most periods end with a remainder, and incidents per period fall. [synthesis]

4.3 Response cost combined with positive reinforcement (earn-and-lose system)

Purpose / when: this is the standard clinical configuration. Tokens are earned for the replacement behaviour and lost for the target behaviour. BACB 2.14 requires that positive reinforcement be prioritised, and Cheney et al. advise always combining response cost with reinforcement and skills teaching. [source: BACB 2022; Cheney et al. 2005]
Steps:
1. Pair each fined behaviour with a functionally equivalent replacement behaviour that earns. Example: grabbing (fine) paired with asking for a turn (earn). [synthesis grounded in BACB 2.14]
2. Design so that typical earnings clearly exceed typical fines. A practical design target is earning opportunities outnumbering fine opportunities. [synthesis]
3. Deliver both promptly and consistently.
4. Review the balance daily with the person, and highlight earnings first.
Example language (illustrative): "Today you earned 14 and lost 3, so you have 11. The earning was mostly from asking for help. That's what's working."
Watch for: staff drifting to fining only. In Iwata & Bailey (1974), the reward condition raised teacher approval comments, but the cost condition produced no change. So monitor adult praise rates as well as fines. [source: Iwata & Bailey 1974 summary]
Working if: the replacement behaviour rises as the target falls. Iwata & Bailey found that reward and cost systems reduced rule violations and off-task behaviour about equally, and students showed no consistent preference between them. [source: Iwata & Bailey 1974 summary]

4.4 Group response cost / group consequences

Purpose / when: a whole group's behaviour, e.g., in a classroom or lunchroom. An example is giving or removing class recess time based on lunchroom behaviour. [source: Cheney et al. 2005; Cooper et al. method list]
Steps:
1. Define the group target behaviours.
2. Give the group a visible shared reserve, such as minutes of recess on the board. [synthesis]
3. Deduct a set amount per defined incident by anyone, announced neutrally without naming the student.
4. Deliver the remainder at the scheduled time.
5. Add group earning for target positive behaviour.
Example language (illustrative): "That was shouting across the room. The class loses one minute. You still have twelve."
Watch for: peer pressure or scapegoating of the student who causes most losses, and a single student deliberately draining the reserve. If either occurs, move that student to an individual contingency. [synthesis] The related Good Behavior Game uses team marks, not removal of held reinforcers. It is a separate procedure and is not covered here. [uncertain]
Working if: group incidents fall, with no rise in peer conflict. [synthesis]

4.5 Time-based cost (loss of free-time or activity minutes)

Purpose / when: the reinforcer is minutes of a preferred activity. Rapport et al. (1982) compared it with two Ritalin doses in two boys with ADHD, using contingent loss of free-time minutes for off-task behaviour. NM DDSD places loss of activity time at the mildest end of its continuum. [source: Rapport et al. 1982 summary; NM DDSD 2024]
Steps:
1. Give a fixed reserve of free-time minutes per work period and display it.
2. Define "off-task" observably.
3. Deduct one unit per observed instance, at the moment it is observed.
4. Deliver the remaining minutes at the end of the period.
5. Pair this with praise for on-task work.
Rapport et al. reportedly used a desk-mounted counter that the teacher decremented remotely. That device detail and the exact minute values were not verified. [uncertain: full paper]
Example language (illustrative): "Eyes on your worksheet, please. That's one minute off." (delivered quietly at the desk or via the display)
Watch for: the loss of free or structured time being inappropriate for some populations, such as people with developmental disabilities, where it may require committee approval. [source: NM DDSD 2024 citing Cooper et al.] Also watch for observer fatigue causing inconsistent deductions. [synthesis]
Working if: on-task behaviour and work completion rise. Rapport et al. reported higher on-task behaviour and academic performance under response cost than under either Ritalin dose. [source: Rapport et al. 1982 summary]

4.6 Contingent partial removal of an item, activity or money

Purpose / when: these are the next steps on NM DDSD's continuum: contingent removal of all or part of a desired item, then contingent loss of money (a fine or repayment programme). Use them when no token system fits, e.g., with adults in supported living, but only with consent and review. [source: NM DDSD 2024]
Steps:
1. Specify the item or amount and the exact portion removed per incident.
2. Obtain informed consent or assent and document that the person understands. Never start it "on the fly". [source: NM DDSD 2024]
3. Plan the handover: how the item or money changes hands without triggering a fight. [source: NM DDSD 2024]
4. Check that the loss cannot create unrecoverable debt.
5. Get human rights committee approval where required, and review it at least quarterly. [source: NM DDSD 2024] Restitution programmes are covered by separate NM guidelines. [source: NM DDSD 2024]
Example language (illustrative): "We agreed that breaking a housemate's things means paying toward replacing them from your weekly spending money. This week that's the amount on your plan. Do you want to hand it over now or after dinner?"
Watch for: secondary aggression during handover, the plan becoming entrenched, and staff using it as an informal threat. [source: NM DDSD 2024]
Working if: fewer incidents, with no new challenging behaviour. Without both, NM DDSD treats the consequence as having no clinical value. [source: NM DDSD 2024]

4.7 Lottery response cost

Purpose / when: a low-effort classroom variant associated with Witt & Elliott (1982). Each student starts the period with a set of slips bearing their name. The teacher removes one slip per rule violation, and the slips left at the end go into a draw for a reward. [uncertain: original paper. Description from secondary memory of the literature; verify before use]
Steps (as generally described):
1. Give each student a fixed number of named slips, kept visible at the desk.
2. Post the rules.
3. Remove one slip per violation, with a one-sentence neutral statement.
4. At the end of the period, collect the remaining slips into a container and draw one or more winners for a modest reward.
5. Optionally add slips for target positive behaviour. [synthesis]
Example language (illustrative): "Talking during the test, Jordan. One slip."
Watch for: students who lose all their slips early have nothing left to lose. Cap losses per period, or give a fresh-start point mid-period. [synthesis] Low-probability rewards may not motivate young children. [synthesis]
Working if: classroom violations fall, and the teacher finds it sustainable. [synthesis]

4.8 Self-managed response cost

Purpose / when: the person records their own target behaviour and removes their own points. This builds self-management. Use it with older children, adolescents or adults who can self-monitor accurately. [source: Cheney et al. 2005]
Steps:
1. Teach the definition with examples and non-examples until the person classifies them accurately. [synthesis]
2. Provide a recording card with the starting points.
3. The person deducts points at each occurrence.
4. The adult also records independently at first. Compare the two records, and give a bonus for matching, whatever the score. [synthesis] Matching checks are common in self-management, but the specific bonus rule is synthesis.
5. Reduce the adult's checks as accuracy holds.
Example language (illustrative): "You noticed you called out and took the point off yourself. That's exactly the skill. And your count matches mine, so that's a bonus point."
Watch for: under-reporting. Keep matching checks going until accuracy is stable. [synthesis]
Working if: self-records agree with independent records, and the behaviour falls. [synthesis]

4.9 In-session contingency in CBT for impulsive children (reward plus response cost)

Purpose / when: the treatment combines self-instruction training with a within-session contingency. It is used in individual child CBT for impulsivity. Kendall & Braswell's Cognitive-Behavioral Therapy for Impulsive Children teaches "stop and think" self-instructional steps through coping modelling. [source: Guilford and Workbook Publishing listings] As commonly described, the child starts each session with a supply of chips or tokens and loses one for specified errors, such as rushing or skipping a step. The child also earns for self-evaluation, and the remaining chips are exchanged for prizes. [uncertain: verify against the manual]
Steps (as generally described):
1. At session start, give the child a set number of chips and explain which errors cost a chip.
2. While you work through the tasks with self-instructions, remove a chip calmly at each listed error, naming it.
3. Model the step that was skipped. [synthesis]
4. At session end, the child rates their own performance. Award bonus chips for an accurate self-rating. [uncertain]
5. Exchange the chips against a prize menu.
Example language (illustrative): "Oops, that one went too fast. That costs a chip. Let's say the steps out loud together: what am I supposed to do?"
Watch for: a perfectionistic or anxious child who reacts strongly to losses. Lower the cost or give a larger starting supply. [synthesis]
Working if: in-session errors fall across sessions, and the child begins using the steps without prompting. [synthesis]

4.10 Fines within a home poker chip or point system (parent-delivered)

Purpose / when: this is used in behavioural parent training for oppositional children. Barkley's Step 4 sets up the home poker chip or point system, and Step 5 adds disciplinary consequences. Fines from the chip system are, as I understand it, used for misbehaviour from Step 5 onward. [source: Barkley 2013 contents; CEBC] [uncertain on exact timing and fine rules]
Steps (therapist teaching parents):
1. In session, help parents list the jobs and behaviours that earn chips or points, with values, plus the reward menu with prices.
2. Have parents run earning only for about a week, giving chips generously. [source: Barkley sequence] The exact duration is uncertain.
3. In the next session, review the chip log. If earning is going well, add fines for a short list of specific misbehaviours, in amounts the parents choose and write down. [uncertain on Barkley's specific fine-size rule]
4. Role-play delivery with the parents: calm, one sentence, immediate.
5. Set a rule that a parent who is angry delays the fine until calm, and never invents a fine. [synthesis]
6. Review the fines/earnings ratio every session.
Example language (illustrative, for parents): "You were asked to turn off the game and you kept playing. That's 5 chips." To the parent in session: "Let's practise saying that in a flat voice. One sentence, then walk away."
Watch for: parents fining much more than they pay, harsh fines at bedtime when nothing can be earned back, and parents fining for one-off annoyances that are not on the list. [synthesis]
Working if: the targeted misbehaviour falls on the parent's log, and the child still earns most days. [synthesis]

4.11 Checkbook or ledger system (token economy with fines paid by "check")

Purpose / when: a secondary-school variant (Siegel 1995, as applied by Smith-Fontenot & Siegel). Students earn "money" for appropriate behaviour, pay fines by writing checks, and keep balanced registers. It suits adolescents and adds maths and self-management practice. [source: Smith-Fontenot & Siegel, ERIC EJ1140164]
Steps:
1. With the students, list the appropriate behaviours (earnings) and inappropriate behaviours (fines), agree the amounts, and post them.
2. Teach check-writing and ledger balancing.
3. Set up a class store with student-chosen items, open at fixed times (e.g., weekly).
4. About 15 minutes before the end of each day, meet each student to settle the day's earnings and fines by check.
5. Optional later additions: rent, random ledger audits with a fine for an unbalanced ledger, tax.
[source: ERIC EJ1140164]
Example language (illustrative): "Your register says 40. Two call-outs today at 3 each, so write me a check for 6. What's your new balance?"
Watch for: the published case shows the risk. The student's target behaviour rose late in the study after rent was added and his money ran out. The authors suggest introducing rent more slowly and requiring savings. [source: ERIC EJ1140164]
Working if: fines fall, and the student maintains a positive balance and uses the store. [synthesis]

4.12 Monetary fines and deposit arrangements with adolescents and adults

Purpose / when: with adults, response cost is mostly self-imposed, e.g., agreed fines from pocket money or a returnable deposit forfeited in part when a goal is missed. Rapport & Begolli describe response cost as loss for an undesirable behaviour or for failing to meet a goal. [source: Rapport & Begolli 2005] Deposit-contract procedures exist in the behavioural literature; no protocol is drawn on here. [uncertain]
Steps (synthesis):
1. The client chooses whether to use a fine and its amount. It must be an amount they can afford to lose without hardship.
2. Define the behaviour or goal precisely and how it will be verified.
3. Agree where forfeited money goes (e.g., a neutral charity), and a cap on total loss.
4. Review the arrangement weekly, and the client can withdraw at any time.
5. Pair it with skill-building and reinforcement for progress.
Example language (illustrative): "You suggested putting some money at stake. Let's make sure it's an amount that would sting but not hurt your budget, and that you can end the arrangement any time."
Watch for: shame, hiding lapses to avoid the fine, and financial harm. Do not use this for eating, weight or exercise goals with anyone at risk of disordered eating. [synthesis]
Working if: the target behaviour moves in the intended direction, and the client reports the arrangement as motivating rather than distressing. [synthesis]

4.13 Monitoring, evaluation and fading routine (applies to every variant)

Purpose / when: this runs throughout. It is required by BACB 2.15, 2.17 and 2.18 and by NM DDSD. [source: BACB 2022; NM DDSD 2024]
Steps:
1. Graph the target behaviour (frequency, severity, duration) against baseline.
2. Graph fines delivered and reinforcers earned.
3. Track collateral measures: aggression, avoidance, desired behaviour, and preferred activities obtained freely.
4. Set a decision rule in advance, e.g., if there is no clear decrease after an agreed trial period, modify or stop. [synthesis]
5. Once the behaviour is stable at a low level, thin fines first: reserve them for more severe instances, or lengthen the review interval. Keep reinforcement in place. [synthesis]
6. Then fade the token system itself towards natural reinforcers. [synthesis]
Example language (to a team, illustrative): "Fines are down to about one a week, and earning is steady. Next month we drop fines for the minor items and keep only the one for hitting."
Watch for: the behaviour returning after fines are removed. If it does, briefly reinstate and fade more gradually. [synthesis] Pfiffner & O'Leary (1987) reported that all-positive classroom management held up better when negative consequences had previously been used and were faded gradually. [uncertain]
Working if: the behaviour stays low with fines rare or absent. [synthesis]

5. Handling problems

Sourced problems are tagged with their source; everything else is [synthesis].

Refusal to hand over tokens or pay the fine.
- Sources: Cooper et al. list avoidance and aggression among the issues with response cost. The U Idaho notes say response cost needs backup consequences for refusal, and that the person must not be able to escape the fine. [source: Cooper et al. via flashcards; U Idaho notes]
- Response: decide the backup in advance and tell the person, e.g., an unpaid fine is deducted at the next exchange, plus a small surcharge, or settled by a brief, previously agreed alternative. Never wrestle for tokens. Use ledgers or electronic counts rather than physical tokens where grabbing is likely. [synthesis]
Debt spiral / balance at zero.
- Sources: when the balance runs out, there is nothing left to lose and motivation collapses. NM DDSD says the programme, not the person, is the point of failure. [source: NM DDSD 2024; U Idaho notes]
- Response: floor the balance at zero, never carry debt forward, and lower fine sizes or the number of fined behaviours. Raise earning opportunities, and use bonus response cost with shorter periods. Do not switch to time-out by default. If you consider it, apply the same least-restrictive analysis. [synthesis]
Escalation or aggression when fined.
- Sources: fines are a major stressor and can escalate behaviour. Secondary aggression is a sign of an unintended effect. [source: NM DDSD 2024; Cooper et al. via flashcards]
- Response: use incident clustering (one fine per episode) and deliver fines quietly. Delay the fine until the person is calm (record it now, deduct later), and plan the handover. If aggression rises on the data, suspend the procedure and reassess function. [source for clustering and handover: NM DDSD 2024; rest synthesis]
No effect.
- Sources: if the behaviour does not decrease, the loss is not functioning as punishment. Reconsider or abandon it; continuing may be abusive in disability services. [source: Cheney et al. 2005; NM DDSD 2024; BACB 2.15]
- Response: check four things before abandoning:
  - Do the tokens actually have value? Check the exchanges.
  - Are fines applied consistently? Check fidelity.
  - Is the fine magnitude meaningful? Set it once, do not ratchet it up.
  - Does the behaviour produce a stronger reinforcer than the fine removes? Check the function. For example, escape from a hard task may be worth losing a point.
  If none of these fixes it, stop. [synthesis, magnitude point from Burchard & Barrera 1972 and NM DDSD]
Collateral drop in desired behaviour; avoidance of the adult.
- Sources: this is listed as a drawback of response cost. [source: Cooper et al. via flashcards]
- Response: raise the praise and earning rate of the adult who fines. Have a different adult deliver the rewards. Review whether the work demands are reasonable. [synthesis]
Staff or parent overuse; fines given in anger or invented on the spot.
- Sources: [source: Cooper et al.; U Idaho notes; NM DDSD 2024]
- Response: allow only listed behaviours and amounts. Audit the fine log against the plan weekly. If fines exceed earnings for more than a couple of days, convene a review. Coach a "fine later, when calm" rule. Retrain with role-play. [synthesis]
Inconsistency between carers or settings.
- Response: give everyone one written plan and one shared log. Do brief fidelity checks, observing a carer deliver a fine. Better to drop the fine in a setting where it cannot be run consistently. [synthesis]
"I don't care about the points" (apparent indifference).
- Response: do not argue. Check the data: if the person still exchanges tokens, the statement is not evidence the procedure is failing. If they truly do not value the tokens, refresh the reward menu with the person, before adding any fines. [synthesis]
Over-reaction in anxious or perfectionistic clients.
- Response: use smaller fines and a larger reserve, or use reinforcement only. Make sure an error can be recovered within the same period. [synthesis]
Clients with trauma histories, or for whom losing possessions is loaded (e.g., care-experienced children).
- Response: discuss the plan in advance, never remove personal or transitional items or contact with family, and prefer reserve-based tokens over real belongings. Monitor distress closely. [synthesis]
Rupture with the client or family.
- Response: name it. Acknowledge that fines feel unfair, revisit the plan together (BACB 2.09 involvement), and adjust amounts or the behaviour list with them. Restore the earning focus. [synthesis; involvement from BACB 2.09]
Non-adherence by parents between sessions.
- Response: shrink the plan to one fined behaviour and keep the earning system. Troubleshoot the specific moments where fining broke down, using the log, and role-play those moments. [synthesis]
Programme entrenchment (it never ends).
- Sources: NM DDSD notes that response cost programmes can last for years. [source: NM DDSD 2024]
- Response: write fading criteria into the original plan, and review them at each quarterly or periodic review (4.13). [synthesis]
Ethical boundary.
- Sources: response cost is a punishment-based procedure. Behaviour analysts must use it only after less intrusive means fail on data, with required review, and must keep evaluating it. [source: BACB 2.15]
- Response: if you are not a behaviour analyst, follow your own profession's code and any local restrictive-practice policy. These may require similar consent and review. [synthesis]

Sources

Sources:

Further reading (not drawn on) [uncertain where relied on]:

Weiner, H. (1962). Some effects of response cost upon human operant behavior. JEAB, 5.
Kazdin, A. E. (1972). Response cost: The removal of conditioned reinforcers for therapeutic change. Behavior Therapy, 3.
Cooper, J. O., Heron, T. E., & Heward, W. L. (2007/2020). Applied Behavior Analysis (2nd/3rd ed.). Pearson.
Barkley, R. A. (2013). Defiant Children: A Clinician's Manual for Assessment and Parent Training (3rd ed.). Guilford.
Kendall, P. C., & Braswell, L. (1985/1993). Cognitive-Behavioral Therapy for Impulsive Children. Guilford.
Witt, J. C., & Elliott, S. N. (1982). The response cost lottery. Journal of School Psychology.
Phillips, E. L. (1968). Achievement Place. JABA, 1.
Pfiffner, L. J., & O'Leary, S. G. (1987). All-positive management and prior negative consequences. JABA, 20.
Lovett, H. (1996). Learning to Listen. Brookes (cited via NM DDSD).

Further reading (not drawn on): the PMC full text of Iwata & Bailey (1974) and its Europe PMC full text.

← Previous · II-033Social-learning milieu programme for schizophrenia Next · II-035 →Time-out