Article Preview
TopIntroduction
The literature on the use of algorithms within the public sector is dense. The use of algorithms is described to improve governance effectiveness and efficiency (Wirtz et al., 2019; Meijer & Grimmelikhuijsen, 2020), to reshape accountability relations and alter how transparency is provided (Peeters, 2020; Giest & Grimmelikhuijsen, 2020), to transform public organisational structures (Lorenz et al., 2020), and to change how street-level decisions are made (Young et al., 2019; Busuioc, 2020). Moreover, algorithms impact all organisational levels; they require choices at the managerial level, as do they affect the work of frontline professionals (Lorenz et al., 2020; Meijer et al., 2021; Bullock et al., 2020).
Authors examining the implementation of algorithms repeatably, but not always explicitly, consider how algorithms are used to build or erode public values. This makes sense, public values are: “citizens collective expectations in respect to government and public services” (Moore, 1995). They are the most important units to assess the performance of public organisations because these values are what constitutes their legitimacy (Jørgensen & Bozeman, 2007). The analysis of a new public policy instrument, such algorithms, therefore always involves illustrating how they affect the creation public values. A direct discussion on how algorithms influence public value creation – a concept that is employed in this article to describe the grand total of all public values that are created by a public organisation or government – is however missing. This is the objective of this article. Using Hood’s (1999) classical framework of three public value clusters, this article makes explicit how algorithms impact the ability of governments and public organisations to create public value. Building off this discussion, which shows that algorithms generate public value trade-offs, guidance is provided on how to manage these dilemmas. In doing so, this article’s contribution to our academic knowledge about algorithms in the public sector is threefold: (1) it present public value as an umbrella for discussing the impact of algorithms on the public sector performance, (2) it describes how the development and use of algorithms in the public sector affects the creation of three different types of public values, and (3) it establishes principles to provide guidance for managing algorithms to create public value.
The information in this paper about the management of algorithms in the public sector is also relevant for practitioners. Algorithms fundamentally change the role that public managers play in the policy and decision making process (Van der Voort et al., 2019). Policy makers are however unprepared for the challenges they face when working with algorithms (Agarwal, 2018). The insights presented in this article can therefore support practitioners in performing this new role.