A group's ability to solve complex problems often depends on its members' ability to learn from each other. We argue that scarcity makes learning difficult because it gives some people an incentive to withhold or misrepresent what they know and other people an incentive to discount or ignore what they hear. Thus, even when a group's collective knowledge is sufficient to a solve a complex problem, scarcity may render its members unwilling or unable to transfer the knowledge that an effective solution requires. Since both scarcity and complexity are present in most group decision making settings, we must develop knowledge of how groups adapt to simultaneous presence of both factors if we want to offer effective remedies for the regular problems they face. We will develop such knowledge by drawing on lessons from both the cognitive and social sciences. We will design mathematical models and conduct laboratory experiments to examine not only how people learn from each other but also how, and whether groups can achieve complex goals in the presence of scarcity. By identifying general conditions under which people can learn from each other, our research provides the foundation for concrete suggestions for how to improve the efficacy of group decision making.