In this thesis, we study several problems at the interface of algorithmic decision-making and society, focusing on the tensions that arise between these processes and social values like fairness and privacy. In the first chapter, we examine the design of financial portfolios which adequately serve all segments of the population. In the second, we examine an allocation setting where the allocator wishes to distribute a scarce resource across many groups fairly, but does not know ahead of time which groups have a need for the resource. In the third, we study a game-theoretic model of information aggregation and the effects of individuals acting to preserve the privacy of their personal beliefs on the collective welfare of the population. Finally, we look at some of the issues that arise from the desire to apply automated techniques to problems in redistricting, including fundamental flaws in the definitions and frameworks typically used.