We implement a one-week partial sleep restriction protocol to investigate the effect of sleep deprivation on joint production in a standard voluntary contributions mechanism (VCM) experiment. Additionally, the effect of sleep restriction on an individual's likelihood of sending costly peer punishment is examined. Actigraphy sleep monitoring watches are used to validate that our random assignment to sleep restricted (SR) and well-rested (WR) conditions generates significant differences in both objective nightly sleep duration and subject sleepiness. Using multiple measures of sleep restriction, and non-parametric as well as regression analysis, we find that when punishment is not available, sleep restriction does not affect the contributions made to joint production. When punishment is available, we find weak evidence that SR subjects contribute more than WR subjects, but there is no evidence that SR and WR subjects differ in the amount they punish others. However, we also find that SR subject contributions are significantly more sensitive to the introduction of peer punishment. SR subject punishment decisions may also be more sensitive to the deviation of their contributions from other group members' contributions and more sensitive to having received punishment themselves. Our results have implications for understanding how the norm enforcement availability may differentially impact individuals depending on their current sleep state.