Uses and effects of data and algorithms in politics and society

3.5. Uses and effects of data and algorithms in politics and societyΒΆ

Looking back on this chapter, we have seen that data and algorithms are a phenomenon of growing reach and importance in politics and society. Digital technology has provided an ever increasing amount of data for an ever growing set of societal fields. Ever more actors start looking at these data riches and try to gain insight or control by making new phenomena, people, and behavior visible. But the translation of these new data into measurements and metrics still remains challenging. Paradoxically, the recent abundance of data has pushed questions of measurement into the background, raising the question of what this new data abundance actually represents.

New data source make new objects visible. This new visibility might allow for the development of new metrics which in turn might allow for greater control and evaluation of social or political processes. But the associated translation process needs to be taken seriously and should not be ignored either in the construction of metrics or their evaluation.

Digital technology has also extended the use and reach of algorithms in society. The growing importance of algorithms in society has raised various questions regarding the ethics of underlying data sets and the effects of use. These questions are discussed in computer science, statistics, and also increasingly the social sciences. Algorithms excel at rolling out standardized procedures, but as we have seen, doing so in scale can go wrong. For the responsible use of algorithms in politics and society we need meaningful transparency about their uses, underlying mechanism, and associated effects.

In this chapter, we have focused on the effect of data and algorithms in two societal fields: political campaigning and journalism. Other areas could have also served as instructive examples. For example, data and algorithms play an important role in predictive policing, finance, or cultural production. There is serious research activity in all these areas, which is creative and might provide instructive templates for the analysis of the role of data and algorithms in fields more closely connected to political science.

Already with our few selected examples, we have seen that different fields react differently to the growing abundance of data or the introduction of algorithms. Importantly, they do not always react in the way we might predict or the popular press might have us belief. Careful studies are needed that focus on technical aspects of measurement and algorithms, process oriented studies looking at the translation of reality into data, metrics, and algorithms, studies looking at how social fields adapt, but also studies looking at the actual effects on individuals subject to algorithmic interventions or the shaping of their option spaces.

The role of data and algorithms is a tangled web of interrelated questions and effects. There are several perspectives needed to untangle it. But untangle it, we must, given the growing availability, use, and importance of data and algorithms in politics and society.