An Examination of Reward Probability and Delivery Delays on Employee Performance

Wine et al. (2019) paid adult office workers with a token-economy system. They changed two things: how long workers waited for the bonus and how often the bonus actually came.

Workers still got monthly paychecks. The test bonus was extra money that could arrive the same day or up to 32 days later. The chance of getting it ranged from 100% down to 10%.

Long waits did not hurt work speed. Output stayed high even when the bonus arrived a month late.

Cutting the chance of getting the bonus to 10% finally slowed the workers. Probability, not delay, drove performance.

Catania et al. (2015) saw the opposite with pigeons. Birds pecked less when grain was delayed even a few seconds. The gap shows humans can tolerate long signaled delays if the reward is money and the context is work.

Richards (1981) helps explain why. Pigeons pecked faster when a light signaled the upcoming delay. Monthly paydays act like that signal for people; the calendar tells them the bonus is still coming.

Wine (2025) ran a close cousin study. He kept the delay short and instead changed the size of the bonus. Even $2.11 per shift lifted filing speed. Together the two Wine papers show both size and probability matter more than delay once the signal is clear.

If you run token economies in adult day programs or sheltered workshops, do not fear weekly or monthly pay cycles. Keep the pay day predictable and well signaled. Instead of shortening the delay, guard the probability. Dropping the chance of reinforcement below 20% is where you will lose responding. Tell your supervisees: a small but sure bonus beats a big maybe.