What is a token economy in ABA?
A token economy is a reinforcement system in which learners earn tokens - stars, points, chips - for target behaviors and exchange them later for chosen backup reinforcers.
Money is the adult token economy: pieces of paper with no inherent value, powerful because of what they exchange for. Clinical token systems borrow the mechanism. Tokens are delivered immediately when the target behavior occurs, then traded on some schedule for backup reinforcers the learner actually wants.
Three design pieces define any token system: what earns a token (the target behaviors, clearly defined), what tokens buy (the backup menu, kept fresh by preference assessment), and the exchange arrangement (how many tokens, how often the store opens). Each is a dial the clinician tunes.
Tokens solve real logistical problems. They bridge delays - the reinforcer at the end of the morning is out of reach, but the token now is not. They keep sessions moving when the true reinforcer is disruptive to deliver mid-task. And because they exchange for many things, they resist satiation better than any single item.
The failure modes are predictable and worth designing against. If tokens accrue but the exchange is stingy, delayed, or stocked with stale items, the tokens lose meaning; if earning is too easy, they inflate. Thinning matters too - well-run systems gradually require more behavior per token and stretch exchange delays so the schedule moves toward what natural environments provide.
The end state to design toward is the token system's own retirement: control passing to natural consequences - praise, finished work, the activity itself - with the tokens faded. A token board that must run forever at first-session density is a signal the transfer step never happened.