Resurgence and downshifts in alternative reinforcement rate

Shahan et al. (2020) asked a simple question: if you cut the rate of alternative reinforcement, how much does the old behavior come back?

They used pigeons in a three-phase lab setup. First, pecking one key paid. Then that key stopped paying and a new key paid at a steady rate. Finally, both keys stopped paying. The team trimmed the payoff on the new key in big or small steps and watched the birds return to the first key.

Bigger cuts in payoff made the old key peck surge back faster and stronger. The jump was not straight-line; it curved up like a hockey stick. A math model called RaC2 tracked the curve almost perfectly.

Ritchey et al. (2023) ran the same downshift with college students pressing buttons for money. The humans showed the same hockey-stick resurgence, proving the rule crosses species.

Falligant et al. (2022) looked back at real FCT cases. When clinicians made large schedule cuts, problem behavior roared back—an exact echo of the pigeon data.

Greer et al. (2024) tested kids with developmental disabilities. They found one twist: gradual early thinning kept resurgence tiny, while sudden big drops sparked the same exponential rebound. The lab rule holds, but pacing the first cuts gives you a shield.

You now have a ruler for extinction bursts. Plan big drops, expect big spikes. Plan small, steady drops, get gentle ripples. Use this when you thin FCT tokens, shorten DRA paydays, or fade edible schedules. Start lean and grow slower; your future self will face fewer surprises.