Fairness in technological design

This paper addresses an important multi-disciplinary issue of current interest, that is, the implications of technological design for fairness. A visual, graphical methodology centered on the Taylor-Russell diagram is proposed to address this issue. The Taylor-Russell diagram helps to identify and explore ways in which predictions built into designs can pit the interests of different constituencies against one another. The configuration of the design represents a trade-off between the interests of the communities involved. Whether or not the trade-off is appropriate constitutes a problem of fairness or distributive justice. The breadth of this methodology is supported by a diversity of examples analyzed. These include a surveillance system, an automotive safety system, a civic information system, and the international food distribution system. These examples provide models for application of the methodology to the analysis of designs in further areas of concern. Limitations of the methodology are also discussed. While it helps to identify and clarify issues of fairness in technology design, the methodology does not provide a general theory of fairness, nor can it provide fair solutions to such issues without appeal to further principles or concepts.