Risk management is a problem humans have faced throughout history and across societies. One way to manage risk is to transfer it to other parties through formal and informal insurance systems. One informal method of self-insurance is limited risk pooling, where individuals can ask for help only when in need. Models suggest that need-based transfer systems may require coordination and common knowledge to be effective. To explore the impact of common knowledge on social coordination and risk pooling in volatile environments, we designed and ran a Risk Pooling Game. We compared participants who played the game with no advance priming or framing to participants who read one of two texts describing real-world systems of risk pooling. Players in the primed games engaged in more repetitive asking and repetitive giving than those in the control games. Players in the primed games also gave more in response to requests and were more likely to respond positively to requests than players in the control games. In addition, players in the primed games were more tolerant of wide differences between what the two players gave and received. These results suggest that the priming texts led players to pay less attention to debt and repayment and more attention to the survival of the other player, and thus to more risk pooling. These results are consistent with findings from fieldwork in small-scale societies that suggest that humans use need-based transfer systems to pool risk when environmental volatility leads to needs with unpredictable timing. Models suggest that the need-based transfer strategy observed in this experiment can outperform debt-based strategies. The results of the present study suggest that the suite of behaviors associated with need-based transfers is an easily triggered part of the human behavioral repertoire.