Use response cost — losing tokens for target behavior failures — with care

Removing a token after a missed behavior can increase compliance, but creates emotional side effects that pure positive systems avoid.

Why it works

Response cost is a negative punishment procedure: a token or earned privilege is removed after the target behavior fails to occur. It increases motivation to avoid the loss (consistent with loss aversion) but also produces negative emotional states, relationship friction with the person administering the system, and potential avoidance of the monitoring context. Azrin’s token economy research established that pure positive systems generally produce less emotional disruption and more durable behavior change than punishment-based ones.

How to do it

1. Use response cost only for situations where positive reinforcement alone is insufficient and where the stakes justify the emotional costs.
2. Keep the cost small relative to the potential earnings — a system where a miss erases all tokens is experienced as a punishment system, not an incentive system.
3. Be consistent: applying response cost only sometimes is more disruptive than either always applying it or never applying it.
4. Monitor for emotional blowback — increased frustration, avoidance, or resentment are signals that the cost structure is too punitive.

Evidence

Response cost is a well-studied operant procedure; its effects on compliance are established, and its emotional side effects are consistently observed relative to positive-only systems. (clinical)

Most token economy research uses hybrid systems; isolating response-cost effects is methodologically difficult. Side effects are less studied in naturalistic adult self-change contexts than in institutional settings.

Common mistake

Designing a system that is primarily loss-based ("I lose tokens for missing") rather than gain-based ("I earn tokens for doing") — loss framing produces anxiety about the monitoring system itself.