Ethics will move from a peripheral concern to a central concern and therefore will become a topic of research.
In the digital age, ethics—more than cost—will become a dominant constraint on researchers. That is, in the future, we will struggle less with what can be done and more with what should be done. As that happens, I expect that the rules-based approach of social scientists and the ad-hoc approach of data scientists will evolve toward something like the principles-based approached described in Chapter 6. I also expect that as ethics becomes a more prominent research constraint, it will become a topic of research itself. In much the same way that social researchers now devote time and energy to developing new methods that enable cheaper and more accurate estimates, I expect that we will also work to develop methods that are more ethically responsible. This change will happen not just because researchers care about ethics as an end, but also because researchers care about ethics as a means to conducting social research.
An example of this trend is the research on differential privacy (Dwork 2008). Imagine that, for example, a hospital has detailed health records and researchers want to understand the patterns in that data. Differentially private algorithms enable people to query the health records to learn about aggregate patterns (e.g., people who smoke are more likely to have cancer) while minimizing the risk of learning anything about the characteristics of any particular individual. Developing these kind of privacy-preserving algorithms has become an active area of research; see Dwork and Roth (2014) for a book-length treatment. Differential privacy is an example of the research community taking an ethical challenge, turning it into a research project, and then making progress on it. This is a pattern that I think we will increasingly see in other areas of social research.
As the power of researchers, often in collaboration with companies and governments, continues to increase, it will become increasingly difficult to avoid complex ethical issues. It has been my experience that many social scientists and data scientists view these ethical issues as a swamp to be avoided. But, I think that avoidance will become increasingly untenable as a strategy. We, as a community, can only address these problems if we jump in and tackle them with the creativity and effort that we apply to other research problems.