Effects of negative punishment contingencies on cocaine self‐administration by rhesus monkeys

Although punishment contingencies are widely used with human drug users, basic research on the effectiveness of these procedures is limited. The present study evaluated the effects of a negative punishment contingency, response‐contingent timeout (TO) presentation, on cocaine‐maintained responding. Rhesus monkeys were trained under a multiple fixed interval (FI) 5‐min cocaine, conjoint FI 5‐min cocaine‐variable‐interval (VI) 30‐sec TO schedule. TO values were either 0 (baseline), 10, 30, or 60 s in length. During the TO periods, the FI clock continued to operate but the discriminative stimuli signaling cocaine availability were removed, and responding had no scheduled consequence. Cocaine maintained responding in all monkeys and the dose–effect curve was characterized as an inverted U‐shaped function. The response‐contingent TO presentations reduced response rates maintained by cocaine in all monkeys compared to baseline. The magnitude of the reduction in response rates was not a function of the length of the TO period (i.e. intensity of the punisher), and the punishment effect was enhanced by increases in cocaine dose. When responding was punished, response rates in the unpunished components either also decreased (i.e. response induction; ∼30% of the cases) or were not affected (∼60%). These results demonstrate that cocaine‐maintained behavior can be decreased by environmental manipulations involving negative punishment contingencies.