A token economy is primarily a consequence-based intervention, not an antecedent intervention. Its core mechanism works by delivering tokens after a desired behavior occurs, making it a reinforcement system. However, the full picture is more nuanced than a simple either/or classification, because tokens also serve antecedent functions that influence future behavior.
How a Token Economy Works
A token economy has three essential components: clearly defined target behaviors, a system for issuing tokens, and a structure for exchanging those tokens for actual rewards (called backup reinforcers). When someone performs the target behavior, they receive a token. Once they’ve accumulated enough tokens, they trade them in for something they want, like free time, a snack, or a preferred activity.
The formal definition captures this sequence clearly: a token economy is a reinforcement system in which appropriate behavior produces secondary reinforcement in the form of tokens that can periodically be exchanged for other reinforcers. The tokens themselves aren’t inherently rewarding. They gain their value from being repeatedly paired with the backup reinforcers, much like money has no inherent value but represents what you can buy with it.
Because the token is delivered contingent on a behavior that has already happened, the primary mechanism is consequence-based. In the Antecedent-Behavior-Consequence (ABC) model, the token sits squarely in the “C” position. The child does the target behavior, then receives the token. That’s reinforcement.
Why Tokens Also Have Antecedent Properties
Here’s where it gets interesting. While earning a token is a consequence for the behavior that just happened, the presence of that token on a token board becomes an antecedent for the next behavior. Tokens acquire what researchers call discriminative functions, meaning they signal that reinforcement is getting closer and prompt continued effort.
A comprehensive review in the Journal of the Experimental Analysis of Behavior documented this effect directly. In one study, when researchers gave participants free tokens at the start of a session, response rates increased even though the total number of responses required to reach the reward hadn’t changed. The tokens on the board created stimulus conditions that signaled the exchange period was approaching, which strengthened motivation. In another experiment, when tokens were removed from view but all other contingencies stayed the same, token-earning behavior decreased significantly. The tokens weren’t just rewards sitting on a board. They were actively cueing future behavior.
Think of it this way: a child looks at a token board with two out of three spots filled and thinks, “I’m almost there.” That visual cue functions as an antecedent for the next target behavior. The tokens are telling the child something about what’s coming, not just rewarding what already happened.
Where It Fits in the ABC Model
In applied behavior analysis, antecedent interventions are strategies that modify what happens before a behavior to make the desired behavior more likely. Common examples include visual schedules, priming, offering choices, and modifying the environment. These strategies work by changing the conditions that set the stage for behavior.
Consequence-based interventions, by contrast, modify what happens after a behavior to strengthen or weaken it. Reinforcement systems, including token economies, fall into this category. The token is earned because of the behavior, and its purpose is to increase the likelihood that the behavior happens again.
A token economy spans both sides of this divide, but its classification as a reinforcement system reflects its primary function. The ABC model doesn’t require interventions to fit neatly into one box. A single system can operate on behavior from multiple directions. The token economy’s defining feature, delivering tokens after target behavior, is a consequence. Its secondary feature, accumulated tokens signaling proximity to a reward, is an antecedent effect that emerges naturally from the system.
Some practitioners deliberately build antecedent elements into a token economy. For instance, one study compared offering a choice of backup reinforcer before the task (an antecedent arrangement) versus after earning all tokens (a consequence arrangement). In the antecedent condition, a picture of the chosen reward was visible on the token board while the child worked. Both approaches used the same token system, but the timing of the choice shifted whether that element functioned as an antecedent or a consequence.
How Effective Token Economies Are
Regardless of classification, token economies produce measurable behavior change. In a study of 50 children with behavioral difficulties, mean behavior scores dropped from 34.4 to 16.3 after a token economy intervention. Before the program, 76% of children had moderate behavioral difficulties and 24% had severe difficulties. Afterward, 36% exhibited normal behavior and the remaining 64% showed only mild difficulties. No children remained in the moderate or severe categories.
These systems are used across a wide range of settings: classrooms, psychiatric units, group homes, and family households. Their flexibility is part of what makes them effective. The tokens can be anything (stickers, points, poker chips), the backup reinforcers can be tailored to individual preferences, and the exchange ratios can be adjusted as behavior improves. Response cost, where tokens are removed for problem behavior, adds a punishment component that further extends the system beyond simple reinforcement.
The Short Answer for ABA Students
If you’re studying for the BCBA exam or an ABA coursework quiz, classify a token economy as a consequence-based reinforcement system. That reflects its primary mechanism and matches how major textbooks define it. But if an exam question asks whether tokens can serve antecedent functions, the answer is also yes. Accumulated tokens act as discriminative stimuli that signal reinforcement availability and influence ongoing behavior. The system is built on consequences, but it generates antecedent effects as a natural byproduct of how tokens accumulate and signal progress toward a reward.