Automated, administrative decision‐making and good governance: Synergies, trade‐offs, and limits

Published date01 November 2024
AuthorUlrik B. U. Roehl,Morten Balle Hansen
Date01 November 2024
DOIhttp://doi.org/10.1111/puar.13799
SCHOLARLY TAKES
Automated, administrative decision-making and good
governance: Synergies, trade-offs, and limits
Ulrik B. U. Roehl
1
| Morten Balle Hansen
2
1
Department of Digitalization, Copenhagen
Business School, Frederiksberg, Denmark
2
Department of Political Science and Public
Management, University of Southern Denmark,
Odense, Denmark
Correspondence
Ulrik B. U. Roehl, Department of Digitalization,
Copenhagen Business School, 60 Howitvej, DK-
2000 Frederiksberg, Denmark.
Email: ubur.digi@cbs.dk
Abstract
Automated, administrative decision-making (AADM) is a key component in digital gov-
ernment reforms. It represents an aspiration for a better and more efficient administra-
tion but also presents challenges to values of public administration. We systematically
review the emerging literature on use of AADM from the perspective of good gover-
nance. Recognizing the inherent tensions of values of public administration, the broad
review identifies key synergies, trade-offs, and limits of AADM and good governance
associated with nine values: Accountability, efficiency, equality, fairness, resilience,
responsiveness, right-to-privacy, rule-of-law, and transparency. While synergies repre-
sent low-hanging fruits, trade-offs and limits are hard casesrepresenting challenges
to good governance. Taking the specific decision-making context into account, practi-
tioners and scholars should attempt to nurture the fruitsand lessen the tensions of
the hard-casesthereby increasing the desirable societal outcomes of use of AADM.
Evidence for practice
Public authorities employing automated, administrative decision-making as well
as relevant regulators must be aware of the inherent synergies, trade-offs, and
limits vis-à-vis good governance.
Six synergies of automated, administrative decision-making and good gover-
nance are identified. They represent the advantages of both automated
decision-making and good governance. Such advantages are particularly associ-
ated with values of efficiency, equality, rule-of-law, and transparency.
Eleven trade-offs and limits of automated administrative decision-making and
good governance which must be balanced to support good governance are
identified. Particular attention should be given to values of accountability, fair-
ness, resilience, responsiveness, resilience, and right-to-privacy.
INTRODUCTION
Half a century ago, Herbert Simon (1973, 269) predicted
that people will be more and more an observer, mod-
erator, maintenance man, and repairman for a nearly
autonomous process that can carry on for significant
intervals of time without direct human intervention.
Simons prediction has since been confirmed by
increasing reliance on information and communication
technology (ICT) in public administration and in society at
large across the globe.
One trend is the accelerating use of automated deci-
sion systems that support and increasingly make formal
administrative decisions (sometimes labeled algorithmic
decisionsor smart decisions). Such automated
decision-making has been praised as a solution to several
problems of public administration (e.g., Kuziemski &
Misuraca, 2020), but has also been criticized for creating
new problems (e.g., Eubanks, 2018).
This article conducts a systematic, critical-integrative
review (George et al., 2023) of the literature concerning
[Correction added on 18 November 2024, after first online publication: the
supplemental video has been added.]
Received: 13 June 2022 Revised: 31 January 2024 Accepted: 12 February 2024
DOI: 10.1111/puar.13799
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited and is not used for commercial purposes.
© 2024 The Authors. Public Administration Review published by Wiley Periodicals LLC on behalf of American Society for Public Administration.
1184 Public Admin Rev. 2024;84:11841199.
wileyonlinelibrary.com/journal/puar
use of automated,administrative decision-making (AADM)
through the prism of good governance.
Administrative decisions are central to public adminis-
tration as it is through those legally binding decisions that
public administrative bodies decide on what is lawful or
not in specific cases in relation to individuals or firms (col-
lectively labeled addressees)
1,2
. Administrative decisions
cover a wide spectrum of decisions both beneficial to
addressees (e.g., granting unemployment benefits or
childcare benefits) and restrictive (e.g., denial of permis-
sion to build a house or denial of parole).
We define AADM as administrative decision-making
that is partly or fully based on automated outputs gener-
ated by algorithmic systems that incorporate relevant
regulations of a given policy area. Such decision-making
ranges on a continuum from semi-automated (the final
decision being made by a public servant based on assis-
tance from the system) to fully automated (the final deci-
sion being made by the system). In practice, the use of
AADM often combines semi and fully-automated
instances and is based on combinations of simple and
advanced technologies. (Peeters, 2020; Roehl, 2022).
Administrative decision-making is a very tangible
manifestation of public authority while values of good
governance provide an important basis for the legitimacy
of this authority. Assessing the pros and cons of its
increasing automation through values of good gover-
nance therefore represents an important and meaningful
basis for an informed understanding of its automation.
Drawing on the voices of Francis Fukuyama, Christopher
Hood, Bo Rothstein, and Robert I. Rotberg, we elaborate an
understanding of good governance as an institutional and
administrative order of government characterized by values
of fairness, rule-of-law, accountability, efficiency, and resilience.
Good governance implies tensions between different
public values (Perry et al., 2014). Hood (1991), for exam-
ple, stresses the inherent tensions between efficiency,
fairness, and resilience while Pollitt and Bouckaert (2017)
emphasize the trade-offs and paradoxes that contradic-
tory values tend to produce. To analyze the relations and
tensions of AADM and good governance, we introduce
the concepts of synergies, trade-offs, and limits. A syner-
getic relation describes how the use of AADM may
enhance the values of good governance. A trade-off
describes how use of AADM may diminish the values of
good governance. Limits describe situations where the
use of AADM beyond a certain threshold or limit substan-
tially changes the relation to values of good governance.
Scholars predict that the use of AADM will increase in
both scope and depth due to advances in, for example,
data processing capacity and machine learning tech-
niques (Juell-Skielse et al., 2022). Empirical examples of
use of AADM are widespread: The fully automated, now
defunct, OCI debt recovery system in Australia is a high-
profile example (Carney, 2018) while a more mundane
one is the semi and fully-automated decision system used
by the Swedish Transportation Agency to assess driver
license applications (Andersson et al., 2018).
The increasing use of AADM is mirrored by a growing
academic literature. Despite advances, research is special-
ized and scattered across disciplines (Peeters, 2020), and
themes and findings tend to diverge. Some scholars
implicitly touch upon issues of good governance; even so,
their contributions primarily focus on singular advantages
or disadvantages of AADM and not on understandings of
good governance. Based on a review of existing literature,
this article explicitly uncovers tensions between increasing
use of AADM and values of good governance, by addres-
sing the following research question: How do the use of
AADM and good governance relate,and to what extent do
the relations represent synergies,trade-offs or limits?
Employing core understandings from the public admin-
istration literature, the article builds on a systematic review
of literature within social sciences including specialized sub-
disciplines such as Critical algorithm studies, eGovernment,
and Information systems. We synthesize the literature by
identifying key synergies, trade-offs, and limits between the
use of AADM and nine key values of good governance.
Providing a holistic review of the discussion of the use
of AADM, the article not only contributes with a broad
overview of an emerging literature but also points to nec-
essary balances between the use of AADM and different
values of good governance. The review thus not only pro-
vides impetus for further research but also direct public
administration practitioners to particular attention points
when employing AADM.
GOOD GOVERNANCE AND AUTOMATED,
ADMINISTRATIVE DECISION-MAKING
The subject of good governance is both an ancient one
(Griffith & Ferrari, 2000), an influential and criticized
reform agenda promoted by the World Bank
(Grindle, 2004), and an academic debate concerning what
constitutes good governance (Perry et al., 2014) and how
it relates to public value (Moore, 1997). The purpose here
is not to review this substantial literature, but to extract
broadly accepted values of good governance that are par-
ticularly relevant to use of AADM.
The literature contains both lengthy and short sugges-
tions of values ranging from 73 public values (Jørgensen &
Bozeman, 2007)toattemptstoreducegoodgovernance
to a few prime values (e.g., Fukuyama, 2013). We attempt
to strike a middle ground. Firstly, a list of values should not
be too long to make the analysis infeasible but include
those values that are particularly relevant to the use of
AADM and the exercise of public authority it entails. For
instance, in his seminal article, Rosenbloom (1983)shows
that public administration operates in a tension between
three approaches: Legality, efficiency, and political feasibil-
ity. Due to the definition of AADM in the above, political
feasibility is excluded from our list.Secondly, the list should
recognize the complexity of public administration and
include contradictory values that may produce trade-offs
and paradoxes (Pollitt & Bouckaert, 2017).
PUBLIC ADMINISTRATION REVIEW 1185

Get this document and AI-powered insights with a free trial of vLex and Vincent AI

Get Started for Free

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex