A concurrent schedule of reinforcement is an arrangement where two or more reinforcement schedules operate at the same time, each tied to a different behavior or response option. The organism, whether a pigeon in a lab or a person in everyday life, can freely switch between the available options. This setup is the primary tool behavioral scientists use to study choice, because it mirrors how the real world works: you almost always have more than one thing you could be doing, and each option comes with its own pattern of rewards.
How a Concurrent Schedule Works
The classic laboratory setup involves two response options available simultaneously. A pigeon, for example, might face two keys it can peck, each connected to its own independent schedule of reinforcement. One key might deliver food on a variable-interval schedule (rewarding the first peck after an unpredictable amount of time passes), while the other key runs on a different variable-interval schedule with its own timing. The animal is free to peck either key at any time, allocating its behavior however it “chooses.”
The key word is independent. Each schedule runs on its own clock regardless of what the organism is doing. If you’re responding on Option A, the timer on Option B keeps ticking. This independence is what makes concurrent schedules so useful for studying preference: the organism’s distribution of behavior across the options reveals something meaningful about how reinforcement drives choice.
Most concurrent schedule research uses variable-interval schedules rather than variable-ratio schedules for a practical reason. Variable-interval schedules make the rate of available reinforcement relatively constant regardless of how fast the organism responds, which isolates the effect of reinforcement frequency on choice. When researchers have tested concurrent variable-interval and variable-ratio pairings, they’ve found that organisms tend to respond more on the variable-interval option, which complicates interpretation.
The Matching Law
The most important finding to come out of concurrent schedule research is the matching law, first described by Richard Herrnstein in 1961. The principle is straightforward: the proportion of behavior an organism directs toward one option matches the proportion of reinforcement that option provides. If Option A delivers twice as many reinforcers as Option B, the organism will respond roughly twice as often on Option A.
This works with time as well as response counts. The proportion of time spent on a behavior matches the proportion of reinforcement it produces. So if you’re in a conversation group and one person gives you approval three times as often as another, you’ll spend roughly three times as long talking to the first person. Researchers confirmed exactly this in a classic study where confederates in a discussion group delivered verbal approval on controlled schedules. Participants allocated their conversation time in direct proportion to the approval each person provided, without any awareness they were doing so.
Perfect matching is an idealized prediction. In practice, behavior often deviates slightly. To handle this, Baum proposed the generalized matching equation in 1974, which adds two parameters. The first, called sensitivity, captures how strongly the organism’s behavior tracks differences in reinforcement. A sensitivity value of 1.0 means perfect matching. Values below 1.0 indicate “undermatching,” where the organism doesn’t shift its behavior as much as reinforcement ratios would predict (a common finding, often because switching between options is easy). Values above 1.0 indicate “overmatching,” where the organism exaggerates its preference for the richer option. The second parameter, bias, captures any consistent preference for one option that isn’t explained by reinforcement rates, perhaps because one response is physically easier or one location is closer.
The Changeover Delay
One important procedural detail in concurrent schedule research is the changeover delay. This is a brief pause, typically a few seconds, imposed after the organism switches from one option to the other. During this delay, no reinforcement is available. Without a changeover delay, organisms tend to rapidly alternate between options, and whatever reinforcer happens to land right after a switch effectively reinforces the act of switching itself rather than sustained responding on either option.
Research has shown that as the changeover delay gets longer (tested from 0 up to 20 seconds), organisms switch less frequently between options. The changeover delay forces more sustained engagement with each choice, producing cleaner data about true preference. It also more closely mirrors real life, where switching between activities usually has a cost in time or effort.
What Shapes Choice Beyond Frequency
Reinforcement rate is the most studied variable in concurrent schedules, but it isn’t the only thing that drives choice. The size or quality of the reinforcer matters too. Given two options that pay off equally often, organisms prefer the one that delivers the bigger or better reward.
Delay is another powerful factor. Reinforcers lose their pull the further into the future they are, a process called delay discounting. This decline follows a specific pattern: value drops steeply at short delays and more gradually at longer ones. Interestingly, the size of the reward interacts with this effect. People are more willing to wait for a larger reward than a smaller one, even when the delay is the same. This is why you might impulsively grab a small snack now rather than wait five minutes for a slightly better snack, but you’d happily wait the same five minutes if the alternative were a full meal.
Effort also plays a role. When one option requires more physical or cognitive work, organisms shift their behavior toward the easier alternative, all else being equal. In real-world terms, this explains why people often choose the path of least resistance even when a harder option offers somewhat better rewards.
Everyday Examples
Concurrent schedules aren’t just a lab abstraction. Nearly every moment of your waking life involves choosing between simultaneously available behaviors, each with its own reinforcement pattern. Scrolling social media versus working on a report is a concurrent schedule: social media delivers small, unpredictable hits of novelty on something like a variable-ratio schedule, while the report delivers reinforcement (a sense of progress, eventual completion) on a much leaner, more delayed schedule. The matching law predicts you’ll allocate more behavior toward whichever option provides richer, more immediate reinforcement, which is exactly why people struggle to put their phones down.
A child in a classroom faces a concurrent schedule too. Paying attention to the teacher might produce occasional praise or good grades (delayed, relatively thin reinforcement), while chatting with a friend produces immediate social reinforcement. The child’s behavior will naturally drift toward whichever option is paying off more richly in the moment.
Clinical Applications
Applied behavior analysts use concurrent schedules strategically when working with children who display problem behaviors. The logic is rooted directly in the matching law: if problem behavior currently “pays” better than appropriate behavior, you can shift the balance by enriching the reinforcement for the appropriate alternative.
In one notable application with children with autism, researchers found that when problem behavior (like aggression or tantrums) produced a break from tasks and task completion produced nothing, problem behavior dominated. That’s exactly what the matching law would predict. But when completing tasks produced a break plus access to preferred activities, while problem behavior still produced only a plain break, the balance flipped. Problem behavior was eliminated and task completion went up dramatically. The treatment gains held even as expectations were gradually increased and reinforcement was thinned out, all without ever using extinction (removing the reinforcer for problem behavior entirely).
This is a significant practical advantage. Extinction procedures can produce temporary bursts of escalated problem behavior, which can be dangerous or impractical in real-world settings. Concurrent schedule arrangements sidestep that problem by making the appropriate behavior the better deal rather than making the problem behavior stop paying off altogether. The organism simply reallocates its behavior toward the richer option, just as the matching law predicts.