Some scholars, looking broadly at the impact of information and communications technologies (ICTs) on public administration, have argued that the use of ICT has led to structural changes in some public agencies, with IT system experts and designers now being seen as the new bureaucrats with discretionary powers (Bovens and Zouridis, 2002). Further, advances in artificial intelligence (AI), machine learning (ML), predictive analytics on big data, and related technologies over the last decade have led to more attention being focused on the use of predictive algorithms for decision-making in the public sector. This trend can fundamentally change how public services are delivered, impact the behavior of both citizens and employees in public organizations, and have a profound impact at a broader political level. We present some findings from recent literature on the subject.
Which decisions are being facilitated by these technologies in the public sector?
Public agencies are increasingly using predictive algorithms to make decisions about citizens’ lives, enforce laws, and assist in regulation-making (Saxena et al., 2021; Bell, 2021). For example, big data is being used for data-driven regulation in the financial sector in the EU (Kempeneer, 2021). Other use cases that researchers have highlighted include the use of AI to profile unemployed individuals and reduce service delivery costs (Desiere and Struyven, 2021). In essence, technologies such as AI are being used in public sector decision-making to detect patterns, sort populations, and make predictions (Henman, 2020).
What are the current and potential applications for these technologies in the public sector?
Predictive algorithms are being used in areas such as child protection services, resource allocation planning (for example, homeless shelter planning), predictive policing, adjudication, and managing government-citizen communication at a large scale (for example, the chatbot used by the Australian Taxation Office) (Henman, 2020; Pi, 2021). AI is also being used in research related to climate change and environmental monitoring (Galaz et al., 2021).
Several scholars have also showcased future applications of predictive algorithms in the public sector. For example, hybrid machine learning models can be used to evaluate urban flood risk and predict the demand for public buses (Rafiei-Sardooi et al., 2021; Bakdur et al., 2021). AI is seen as a tool for health practitioner regulation, as has been suggested in the Australian context (Wolf, 2020).
What do public agencies need to watch out for when adopting these technologies?
Existing scholarship generally tends to take a rational view of these technologies in the public sector. However, these technologies cannot be isolated from politics; they need to be seen within appropriate environmental and organizational contexts. One particular study used findings from three case studies to argue that AI adoption in government is contingent not so much on technical factors such as data quality but on environmental and organizational factors. The authors identified local pressure, networks, private vendors, isomorphism, and regulation as the environmental antecedents, and organizational funding, IT resources, end-user participation, training, management support, organizational culture, and incentives as the organizational antecedents (van Noordt and Misuraca, 2020). Challenges for government in AI implementation can be categorized into 1) social challenges, 2) economic challenges, 3) data challenges, 4) organizational and managerial challenges, 5) technological and technology implementation challenges, 6) political, legal, and policy challenges, and 7) ethical challenges (Dwivedi et al., 2021).
An important facet of the adoption of predictive algorithms in the public sector is the impact on bureaucracy and the larger organization. Scholars have argued that artificial bureaucrats have disrupted the traditional control mechanisms and suggest that AI agents and human agents must co-work (Bullock and Kim, 2020). Other scholars have examined the historical context behind big data adoption and concluded that the push for such practices can be attributed to “a distinct political culture, a representative democracy undermined by pervasive public distrust and uncertainty” (Rieder and Simon, 2016). Another study has questioned the impact of big data and algorithmic systems on decision-making – an analysis of two case studies found that big data cannot be isolated from managerial and political institutions, and big data too provides opportunities for pursuing self-interest (van der Voort et al., 2019).
Finally, while predictive algorithms are already finding use in many public sector areas, recent novel applications suggest that the use of predictive technologies in the public sector is at a nascent stage, with much scope for future applications. However, successful adoption of these technologies in the public sector is contingent on a number of non-technical factors. If not designed and implemented carefully, these technologies can not only end up as gimmicks but generate significant harms for both citizens and public organizations.
Bakdur, A., Masui, F., & Ptaszynski, M. (2021). Predicting Increase in Demand for Public Buses in University Students Daily Life Needs: Case Study Based on a City in Japan. Sustainability, 13(9), 5137.
Bell, B. W. (2021). Replacing Bureaucrats with Automated Sorcerers?. Dædalus, 150(3).
Bovens, M., & Zouridis, S. (2002). From street‐level to system‐level bureaucracies: how information and communication technology is transforming administrative discretion and constitutional control. Public administration review, 62(2), 174-184.
Bullock, J. B., & Kim, K. C. (2020). Creation of artificial bureaucrats. In Proceedings of European Conference on the Impact of Artificial Intelligence and Robotics 2020.
Desiere, S., & Struyven, L. (2021). Using artificial intelligence to classify jobseekers: the accuracy-equity trade-off. Journal of Social Policy, 50(2), 367-385.
Dwivedi, Y. K., Hughes, L., Ismagilova, E., Aarts, G., Coombs, C., Crick, T., … & Williams, M. D. (2021). Artificial Intelligence (AI): Multidisciplinary perspectives on emerging challenges, opportunities, and agenda for research, practice and policy. International Journal of Information Management, 57, 101994.
Galaz, V., Centeno, M. A., Callahan, P. W., Causevic, A., Patterson, T., Brass, I., … & Levy, K. (2021). Artificial intelligence, systemic risks, and sustainability. Technology in Society, 67, 101741.
Henman, P. (2020). Improving public services using artificial intelligence: possibilities, pitfalls, governance. Asia Pacific Journal of Public Administration, 42(4), 209-221.
Kempeneer, S. (2021). A big data state of mind: Epistemological challenges to accountability and transparency in data-driven regulation. Government Information Quarterly, 101578.
Pi, Y. (2021). Machine learning in Governments: Benefits, Challenges and Future Directions. JeDEM-eJournal of eDemocracy and Open Government, 13(1), 203-219.
Rafiei-Sardooi, E., Azareh, A., Choubin, B., Mosavi, A. H., & Clague, J. J. (2021). Evaluating urban flood risk using hybrid method of TOPSIS and machine learning. International Journal of Disaster Risk Reduction, 66, 102614.
Rieder, G., & Simon, J. (2016). Datatrust: Or, the political quest for numerical evidence and the epistemologies of Big Data. Big Data & Society, 3(1), 2053951716649398.
Saxena, D., Badillo-Urquiola, K., Wisniewski, P. J., & Guha, S. (2021). A framework of high-stakes algorithmic decision-making for the public sector developed through a case study of child-welfare. Proceedings of the ACM on Human-Computer Interaction, 5(CSCW2), 1-41
van der Voort, H. G., Klievink, A. J., Arnaboldi, M., & Meijer, A. J. (2019). Rationality and politics of algorithms. Will the promise of big data survive the dynamics of public decision making?. Government Information Quarterly, 36(1), 27-38.
van Noordt, C., & Misuraca, G. (2020). Exploratory insights on artificial intelligence for government in Europe. Social Science Computer Review, 0894439320980449.
Wolf, G. (2020). Embracing the Future: Using Artificial Intelligence in Australian Health Practitioner Regulation. Journal of law and Medicine, 28(1), 21-44.