Pub Date : 2025-03-01Epub Date: 2025-02-08DOI: 10.1016/j.giq.2025.102010
Hyacinth Balediata Bangero
Although Facebook is seen as a powerful and low-cost tool, insufficient manpower, time, budget, and technical skills hinder effective local government use. Citizens value government pronouncements directly affecting them, especially during uncertain times when guidelines keep changing and are unique per locality. Thus, the study sought the social media use of the 25 most successful cities' official Facebook pages to reveal best practices in e-Government communication for practitioners to learn how to use the relatively new tool efficiently. Using content analysis and anchoring on network analysis theory, the study revealed best practices in posting frequency, post type, shape, length, and topics based on the constructed week sample. Overall, city governments led by younger mayors achieve higher communication success rates. Communication success was also found to be related to the frequency of posting and professionalization. Findings and implications are discussed to help practitioners improve the government's social media utilization.
{"title":"Best practices in e-government communication: Lessons from the local Governments' use of official facebook pages","authors":"Hyacinth Balediata Bangero","doi":"10.1016/j.giq.2025.102010","DOIUrl":"10.1016/j.giq.2025.102010","url":null,"abstract":"<div><div>Although Facebook is seen as a powerful and low-cost tool, insufficient manpower, time, budget, and technical skills hinder effective local government use. Citizens value government pronouncements directly affecting them, especially during uncertain times when guidelines keep changing and are unique per locality. Thus, the study sought the social media use of the 25 most successful cities' official Facebook pages to reveal best practices in e-Government communication for practitioners to learn how to use the relatively new tool efficiently. Using content analysis and anchoring on network analysis theory, the study revealed best practices in posting frequency, post type, shape, length, and topics based on the constructed week sample. Overall, city governments led by younger mayors achieve higher communication success rates. Communication success was also found to be related to the frequency of posting and professionalization. Findings and implications are discussed to help practitioners improve the government's social media utilization.</div></div>","PeriodicalId":48258,"journal":{"name":"Government Information Quarterly","volume":"42 1","pages":"Article 102010"},"PeriodicalIF":7.8,"publicationDate":"2025-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143350163","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-01Epub Date: 2024-12-19DOI: 10.1016/j.giq.2024.101988
Tony Busker , Sunil Choenni , Mortaza S. Bargh
There are many good reasons to use synthetic data instead of real data for research purposes. These reasons may range from the business sensitiveness of real data to increased cost of collecting real data in accordance with GDPR requirements. In this paper, we elaborate upon the potentials of the Large Language Model GPT as a tool to generate synthetic data for analytical purposes when there is no real-data available or accessible. Primarily, we show that by varying the scope of probes adequately, we can generate data of different granularities. To show this, we generated stereotypical data with three levels of granularity by posing more than 18,500 probes to GPT. In total, we generated stereotypical data for eight different views, which can be categorized in three view types corresponding to the three levels of granularity. Secondarily, we show that by varying the scope of probes one can create meaningful information. To show this, we performed a so-called similarity analysis on the generated stereotypical data. We used data visualizations, e.g. heatmaps, to show the views and categories within the views that are similar and those that are at odd with each other. We elaborate upon the application areas of the insight gained about such similarities and differences. Furthermore, we discuss several other types of analysis that can be performed on the generated stereotypical data.
{"title":"Exploiting GPT for synthetic data generation: An empirical study","authors":"Tony Busker , Sunil Choenni , Mortaza S. Bargh","doi":"10.1016/j.giq.2024.101988","DOIUrl":"10.1016/j.giq.2024.101988","url":null,"abstract":"<div><div>There are many good reasons to use synthetic data instead of real data for research purposes. These reasons may range from the business sensitiveness of real data to increased cost of collecting real data in accordance with GDPR requirements. In this paper, we elaborate upon the potentials of the Large Language Model GPT as a tool to generate synthetic data for analytical purposes when there is no real-data available or accessible. Primarily, we show that by varying the scope of probes adequately, we can generate data of different granularities. To show this, we generated stereotypical data with three levels of granularity by posing more than 18,500 probes to GPT. In total, we generated stereotypical data for eight different views, which can be categorized in three view types corresponding to the three levels of granularity. Secondarily, we show that by varying the scope of probes one can create meaningful information. To show this, we performed a so-called similarity analysis on the generated stereotypical data. We used data visualizations, e.g. heatmaps, to show the views and categories within the views that are similar and those that are at odd with each other. We elaborate upon the application areas of the insight gained about such similarities and differences. Furthermore, we discuss several other types of analysis that can be performed on the generated stereotypical data.</div></div>","PeriodicalId":48258,"journal":{"name":"Government Information Quarterly","volume":"42 1","pages":"Article 101988"},"PeriodicalIF":7.8,"publicationDate":"2025-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143136134","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-01Epub Date: 2025-01-20DOI: 10.1016/j.giq.2025.102008
Mary K. Feeney , Federica Fusi , Ignacio Pezo
Open government data (OGD) seeks to promote transparency and accountability by enabling public access to government data. While public managers are increasingly supportive of OGD initiatives worldwide, researchers note that they also carefully select which data to release to balance openness with traditional values of professionalism and secrecy as well as concerns about cyber incidents and privacy. Understanding the factors that influence this micro-level choice is important to make valuable types of data publicly accessible. Using 2018 survey data from a nationally representative sample of 2500 department heads in 500 small and medium-sized US cities, we look at variation in public managers' level of comfort with making different types of government data open - from criminal records to government employee salary data. We find that managerial comfort reflects historic practices of public accessibility and privacy concerns with individual data. Managers who believe OGD creates positive outcomes for society are more comfortable with publicly disclosing all types of data. We also find variation across department types, suggesting fragmented views towards OGD within public organizations.
{"title":"Which data should be publicly accessible? Dispatches from public managers","authors":"Mary K. Feeney , Federica Fusi , Ignacio Pezo","doi":"10.1016/j.giq.2025.102008","DOIUrl":"10.1016/j.giq.2025.102008","url":null,"abstract":"<div><div>Open government data (OGD) seeks to promote transparency and accountability by enabling public access to government data. While public managers are increasingly supportive of OGD initiatives worldwide, researchers note that they also carefully select which data to release to balance openness with traditional values of professionalism and secrecy as well as concerns about cyber incidents and privacy. Understanding the factors that influence this micro-level choice is important to make valuable types of data publicly accessible. Using 2018 survey data from a nationally representative sample of 2500 department heads in 500 small and medium-sized US cities, we look at variation in public managers' level of comfort with making different types of government data open - from criminal records to government employee salary data. We find that managerial comfort reflects historic practices of public accessibility and privacy concerns with individual data. Managers who believe OGD creates positive outcomes for society are more comfortable with publicly disclosing all types of data<em>.</em> We also find variation across department types, suggesting fragmented views towards OGD within public organizations.</div></div>","PeriodicalId":48258,"journal":{"name":"Government Information Quarterly","volume":"42 1","pages":"Article 102008"},"PeriodicalIF":7.8,"publicationDate":"2025-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143136058","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2025-03-01Epub Date: 2025-01-10DOI: 10.1016/j.giq.2024.102006
Helen K. Liu , MuhChyun Tang , Antoine Serge J. Collard
With the increasing attention paid to artificial intelligence (AI) and crowd intelligence (CI) in government, their connections still need to be explored. This study explores the dynamic relationship between AI and CI that constitutes hybrid intelligence for the public sector. Thus, we adopt a bibliometric analysis to identify trends, emerging themes, topics, and interconnections between these two streams of literature. Our review illustrates the intersection between AI and CI, revealing that AI designs can improve efficiency from CI inputs. Meanwhile, AI advancement depends on the quality of CI data. Furthermore, our review highlights key domains such as smart cities (Internet of Things), personnel design, social media, and governance through cases. Based on these illustrated cases, we conceptualize a hybrid intelligence spectrum, ranging from “engagement” to “efficiency,” with crowd intelligence anchoring the former through its emphasis on public participation and AI anchoring the latter through its focus on automation and optimization. Hybrid intelligence, encompassing various forms, occupies the middle ground to balance maximizing public engagement and achieving computational efficiency. Additionally, we elaborate on components of hybrid intelligence designs regarding input (conscious crowds and unconscious crowds), process (algorithmic management and artificial discretion), and outcome (user-focus benefits and non-user-focus outputs). Finally, we recommend prioritizing questions related to the design, regulation, and governance of hybrid intelligence for the public sector.
{"title":"Hybrid intelligence for the public sector: A bibliometric analysis of artificial intelligence and crowd intelligence","authors":"Helen K. Liu , MuhChyun Tang , Antoine Serge J. Collard","doi":"10.1016/j.giq.2024.102006","DOIUrl":"10.1016/j.giq.2024.102006","url":null,"abstract":"<div><div>With the increasing attention paid to artificial intelligence (AI) and crowd intelligence (CI) in government, their connections still need to be explored. This study explores the dynamic relationship between AI and CI that constitutes hybrid intelligence for the public sector. Thus, we adopt a bibliometric analysis to identify trends, emerging themes, topics, and interconnections between these two streams of literature. Our review illustrates the intersection between AI and CI, revealing that AI designs can improve efficiency from CI inputs. Meanwhile, AI advancement depends on the quality of CI data. Furthermore, our review highlights key domains such as smart cities (Internet of Things), personnel design, social media, and governance through cases. Based on these illustrated cases, we conceptualize a hybrid intelligence spectrum, ranging from “engagement” to “efficiency,” with crowd intelligence anchoring the former through its emphasis on public participation and AI anchoring the latter through its focus on automation and optimization. Hybrid intelligence, encompassing various forms, occupies the middle ground to balance maximizing public engagement and achieving computational efficiency. Additionally, we elaborate on components of hybrid intelligence designs regarding input (conscious crowds and unconscious crowds), process (algorithmic management and artificial discretion), and outcome (user-focus benefits and non-user-focus outputs). Finally, we recommend prioritizing questions related to the design, regulation, and governance of hybrid intelligence for the public sector.</div></div>","PeriodicalId":48258,"journal":{"name":"Government Information Quarterly","volume":"42 1","pages":"Article 102006"},"PeriodicalIF":7.8,"publicationDate":"2025-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143136054","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01Epub Date: 2024-10-10DOI: 10.1016/j.giq.2024.101977
Muhammad Afzal , Panos Panagiotopoulos
Research on digital transformation has focused on organizational aspects with less attention to the impacts on frontline public officials' daily work. Drawing on the coping model of user adaptation, we investigate how public officials cope with digitalization initiatives and the role of discretion in the coping process. The empirical study focuses on policing and the responses of law enforcement officials (n = 292) in the Punjab region of Pakistan following the implementation of an integrated Police Station Record Management System (PSRMS). Police officers adopted diverse problem-focused and emotion-focused coping strategies. We examine the relationship between these user adaptation strategies and officers' perceptions of system outcomes in improving performance. The study extends coping theory in the context of frontline digital government interactions and offers guidance on how to better embed systems like the PSRMS in public officials' daily work.
{"title":"Coping with digital transformation in frontline public services: A study of user adaptation in policing","authors":"Muhammad Afzal , Panos Panagiotopoulos","doi":"10.1016/j.giq.2024.101977","DOIUrl":"10.1016/j.giq.2024.101977","url":null,"abstract":"<div><div>Research on digital transformation has focused on organizational aspects with less attention to the impacts on frontline public officials' daily work. Drawing on the coping model of user adaptation, we investigate how public officials cope with digitalization initiatives and the role of discretion in the coping process. The empirical study focuses on policing and the responses of law enforcement officials (<em>n</em> = 292) in the Punjab region of Pakistan following the implementation of an integrated Police Station Record Management System (PSRMS). Police officers adopted diverse problem-focused and emotion-focused coping strategies. We examine the relationship between these user adaptation strategies and officers' perceptions of system outcomes in improving performance. The study extends coping theory in the context of frontline digital government interactions and offers guidance on how to better embed systems like the PSRMS in public officials' daily work.</div></div>","PeriodicalId":48258,"journal":{"name":"Government Information Quarterly","volume":"41 4","pages":"Article 101977"},"PeriodicalIF":7.8,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142421898","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01Epub Date: 2024-11-29DOI: 10.1016/j.giq.2024.101986
Rodrigo Sandoval-Almazan , Adrian Osiel Millan-Vargas , Rigoberto Garcia-Contreras
The implementation of artificial intelligence in the public sector is a fast-evolving tendency in recent years. Despite much research on AI in government- ethics, algorithms, chatbots, AI systems-implement- there is very little data and understanding of the public manager's perception, adaptation, challenges, and resistance on this topic. What are the skills and knowledge needed to implement AI in the government? This research aims to investigate public managers' competencies to face AI challenges in the public sector. A survey was conducted among 38 key public managers from the government of the State of Mexico in the central region to assess their perceptions of AI. Using the competences for civil servants' framework from Balbo di Vinadio et al. (2022), we analyze three competences: (1) Digital Management and Execution (2) Digital Planning and Design (3) Data use and governance and their levels of. The findings point out that there is a lack of skills, and the competence of digital management and execution is the one that explains better this perception of AI in the local government.
近年来，人工智能在公共部门的应用是一个快速发展的趋势。尽管有很多关于政府中人工智能的研究——伦理、算法、聊天机器人、人工智能系统的实施——但很少有关于公共管理者对这个话题的感知、适应、挑战和抵制的数据和理解。在政府推行人工智能所需的技能和知识是什么？本研究旨在调查公共部门管理者应对人工智能挑战的能力。对中部地区墨西哥政府的38名主要公共管理人员进行了一项调查，以评估他们对人工智能的看法。使用Balbo di Vinadio等人（2022）的公务员能力框架，我们分析了三种能力：(1)数字管理和执行(2)数字规划和设计(3)数据使用和治理及其水平。调查结果指出，缺乏技能，数字管理和执行的能力可以更好地解释地方政府对人工智能的这种看法。
{"title":"Examining public managers' competencies of artificial intelligence implementation in local government: A quantitative study","authors":"Rodrigo Sandoval-Almazan , Adrian Osiel Millan-Vargas , Rigoberto Garcia-Contreras","doi":"10.1016/j.giq.2024.101986","DOIUrl":"10.1016/j.giq.2024.101986","url":null,"abstract":"<div><div>The implementation of artificial intelligence in the public sector is a fast-evolving tendency in recent years. Despite much research on AI in government- ethics, algorithms, chatbots, AI systems-implement- there is very little data and understanding of the public manager's perception, adaptation, challenges, and resistance on this topic. What are the skills and knowledge needed to implement AI in the government? This research aims to investigate public managers' competencies to face AI challenges in the public sector. A survey was conducted among 38 key public managers from the government of the State of Mexico in the central region to assess their perceptions of AI. Using the competences for civil servants' framework from Balbo di Vinadio et al. (2022), we analyze three competences: (1) Digital Management and Execution (2) Digital Planning and Design (3) Data use and governance and their levels of. The findings point out that there is a lack of skills, and the competence of digital management and execution is the one that explains better this perception of AI in the local government.</div></div>","PeriodicalId":48258,"journal":{"name":"Government Information Quarterly","volume":"41 4","pages":"Article 101986"},"PeriodicalIF":7.8,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142742944","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01Epub Date: 2024-09-25DOI: 10.1016/j.giq.2024.101974
Ida Lindgren
Automation of public service provision has gained renewed attention as emerging technologies are said to enable automation of tasks that were previously seen as requiring human involvement. However, the merits of these automation technologies are often exaggerated. More knowledge is needed on public service automation, and much can be learned from adjacent research fields studying human-automation interaction. To lead by example, this work applies Bainbridge's (1983) concept of ironies of automation. The purpose is to (1) present ironies of automation, (2) explicate how these ironies can come into play when implementing automated systems in the public service context, and (3) outline implications that follow for public service automation. This is achieved by relating ironies of automation to contemporary studies on Robotic Process Automation (RPA) developments in Swedish local government. The analysis results in five ironies and a set of implications for public service automation. The ironies and implications for public service automation direct attention to key challenges that must be acknowledged in future automation implementations and show that further investigations and theoretical developments are needed on e.g., problems introduced by automation; tasks, roles, and responsibilities that follow on automation; how to design the interface between humans and automated systems in a way that facilitates monitoring, take-over, and maintenance; and, tools and methods for assessing the impact and quality of automated systems. This paper thus provides a foundation for future empirical investigations and further theoretical development on public service automation.
{"title":"Ironies of automation and their implications for public service automation","authors":"Ida Lindgren","doi":"10.1016/j.giq.2024.101974","DOIUrl":"10.1016/j.giq.2024.101974","url":null,"abstract":"<div><div>Automation of public service provision has gained renewed attention as emerging technologies are said to enable automation of tasks that were previously seen as requiring human involvement. However, the merits of these automation technologies are often exaggerated. More knowledge is needed on public service automation, and much can be learned from adjacent research fields studying human-automation interaction. To lead by example, this work applies <span><span>Bainbridge's (1983)</span></span> concept of <em>ironies of automation</em>. The purpose is to (1) present ironies of automation, (2) explicate how these ironies can come into play when implementing automated systems in the public service context, and (3) outline implications that follow for public service automation. This is achieved by relating ironies of automation to contemporary studies on Robotic Process Automation (RPA) developments in Swedish local government. The analysis results in five ironies and a set of implications for public service automation. The ironies and implications for public service automation direct attention to key challenges that must be acknowledged in future automation implementations and show that further investigations and theoretical developments are needed on e.g., problems introduced by automation; tasks, roles, and responsibilities that follow on automation; how to design the interface between humans and automated systems in a way that facilitates monitoring, take-over, and maintenance; and, tools and methods for assessing the impact and quality of automated systems. This paper thus provides a foundation for future empirical investigations and further theoretical development on public service automation.</div></div>","PeriodicalId":48258,"journal":{"name":"Government Information Quarterly","volume":"41 4","pages":"Article 101974"},"PeriodicalIF":7.8,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142318994","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2024-12-01Epub Date: 2024-11-23DOI: 10.1016/j.giq.2024.101982
Antonio Cordella , Francesco Gualdi
Existing literature has predominantly concentrated on the legal, ethical, governance, political, and socioeconomic aspects of AI regulation, often relegating the technological dimension to the periphery, reflecting the design, use, and development of AI regulatory frameworks that are technology-neutral. The emergence and widespread use of generative AI models present new challenges for public regulators aiming at implementing effective regulatory interventions. Generative AI operates on distinctive technological properties that require a comprehensive understanding prior to the deployment of pertinent regulation. This paper focuses on the recent case of the suspension of ChatGPT in Italy to explore the impact the specific technological fabric of generative AI has on the effectiveness of technology-neutral regulation. By drawing on the findings of an exploratory case study, this paper contributes to the understanding of the tensions between the specific technological features of generative AI and the effectiveness of a technology-neutral regulatory framework. The paper offers relevant implications to practice arguing that until this tension is effectively addressed, public regulatory interventions are likely to underachieve their intended objectives.
{"title":"Regulating generative AI: The limits of technology-neutral regulatory frameworks. Insights from Italy's intervention on ChatGPT","authors":"Antonio Cordella , Francesco Gualdi","doi":"10.1016/j.giq.2024.101982","DOIUrl":"10.1016/j.giq.2024.101982","url":null,"abstract":"<div><div>Existing literature has predominantly concentrated on the legal, ethical, governance, political, and socioeconomic aspects of AI regulation, often relegating the technological dimension to the periphery, reflecting the design, use, and development of AI regulatory frameworks that are technology-neutral. The emergence and widespread use of generative AI models present new challenges for public regulators aiming at implementing effective regulatory interventions. Generative AI operates on distinctive technological properties that require a comprehensive understanding prior to the deployment of pertinent regulation. This paper focuses on the recent case of the suspension of ChatGPT in Italy to explore the impact the specific technological fabric of generative AI has on the effectiveness of technology-neutral regulation. By drawing on the findings of an exploratory case study, this paper contributes to the understanding of the tensions between the specific technological features of generative AI and the effectiveness of a technology-neutral regulatory framework. The paper offers relevant implications to practice arguing that until this tension is effectively addressed, public regulatory interventions are likely to underachieve their intended objectives.</div></div>","PeriodicalId":48258,"journal":{"name":"Government Information Quarterly","volume":"41 4","pages":"Article 101982"},"PeriodicalIF":7.8,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142705244","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
AI-driven decision-making systems are becoming instrumental in the public sector, with applications spanning areas like criminal justice, social welfare, financial fraud detection, and public health. While these systems offer great potential benefits to institutional decision-making processes, such as improved efficiency and reliability, these systems face the challenge of aligning machine learning (ML) models with the complex realities of public sector decision-making. In this paper, we examine five key challenges where misalignment can occur, including distribution shifts, label bias, the influence of past decision-making on the data side, as well as competing objectives and human-in-the-loop on the model output side. Our findings suggest that standard ML methods often rely on assumptions that do not fully account for these complexities, potentially leading to unreliable and harmful predictions. To address this, we propose a shift in modeling efforts from focusing solely on predictive accuracy to improving decision-making outcomes. We offer guidance for selecting appropriate modeling frameworks, including counterfactual prediction and policy learning, by considering how the model estimand connects to the decision-maker's utility. Additionally, we outline technical methods that address specific challenges within each modeling approach. Finally, we argue for the importance of external input from domain experts and stakeholders to ensure that model assumptions and design choices align with real-world policy objectives, taking a step towards harmonizing AI and public sector objectives.
人工智能驱动的决策系统正在成为公共部门的重要工具，其应用领域涵盖刑事司法、社会福利、金融欺诈检测和公共卫生等。虽然这些系统为机构决策过程提供了巨大的潜在好处，如提高效率和可靠性，但这些系统也面临着将机器学习（ML）模型与公共部门决策的复杂现实相协调的挑战。在本文中，我们研究了可能出现不匹配的五大挑战，包括分布偏移、标签偏差、数据方面过去决策的影响，以及模型输出方面的竞争目标和人为环路。我们的研究结果表明，标准的 ML 方法往往依赖于无法充分考虑这些复杂性的假设，从而可能导致不可靠和有害的预测。为了解决这个问题，我们建议将建模工作从单纯关注预测准确性转向改善决策结果。通过考虑模型估计值与决策者效用之间的联系，我们为选择适当的建模框架（包括反事实预测和政策学习）提供了指导。此外，我们还概述了应对每种建模方法中特定挑战的技术方法。最后，我们论证了来自领域专家和利益相关者的外部意见的重要性，以确保模型假设和设计选择符合现实世界的政策目标，从而朝着协调人工智能和公共部门目标的方向迈出一步。
{"title":"Bridging the gap: Towards an expanded toolkit for AI-driven decision-making in the public sector","authors":"Unai Fischer-Abaigar , Christoph Kern , Noam Barda , Frauke Kreuter","doi":"10.1016/j.giq.2024.101976","DOIUrl":"10.1016/j.giq.2024.101976","url":null,"abstract":"<div><div>AI-driven decision-making systems are becoming instrumental in the public sector, with applications spanning areas like criminal justice, social welfare, financial fraud detection, and public health. While these systems offer great potential benefits to institutional decision-making processes, such as improved efficiency and reliability, these systems face the challenge of aligning machine learning (ML) models with the complex realities of public sector decision-making. In this paper, we examine five key challenges where misalignment can occur, including distribution shifts, label bias, the influence of past decision-making on the data side, as well as competing objectives and human-in-the-loop on the model output side. Our findings suggest that standard ML methods often rely on assumptions that do not fully account for these complexities, potentially leading to unreliable and harmful predictions. To address this, we propose a shift in modeling efforts from focusing solely on predictive accuracy to improving decision-making outcomes. We offer guidance for selecting appropriate modeling frameworks, including counterfactual prediction and policy learning, by considering how the model estimand connects to the decision-maker's utility. Additionally, we outline technical methods that address specific challenges within each modeling approach. Finally, we argue for the importance of external input from domain experts and stakeholders to ensure that model assumptions and design choices align with real-world policy objectives, taking a step towards harmonizing AI and public sector objectives.</div></div>","PeriodicalId":48258,"journal":{"name":"Government Information Quarterly","volume":"41 4","pages":"Article 101976"},"PeriodicalIF":7.8,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142421899","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Amidst concerns over biased and misguided government decisions arrived at through algorithmic treatment, it is important for members of society to be able to perceive that public authorities are making fair, accurate, and trustworthy decisions. Inspired in part by equity and procedural justice theories and by theories of attitudes towards technologies, we posited that the perception of these attributes of decisions is influenced by the type of explanation offered, which can be input-based, group-based, case-based, or counterfactual. We tested our hypotheses with two studies, each of which involved a pre-registered online survey experiment conducted in December 2022. In both studies, the subjects (N = 1200) were officers in high positions at stock companies registered in Japan, who were presented with a scenario consisting of an algorithmic decision made by a public authority: a ministry's decision to reject a grant application from their company (Study 1) and a tax authority's decision to select their company for an on-site tax inspection (Study 2). The studies revealed that offering the subjects some type of explanation had a positive effect on their attitude towards a decision, to various extents, although the detailed results of the two studies are not robust. These findings call for a nuanced inquiry, both in research and practice, into how to best design explanations of algorithmic decisions from societal and human-centric perspectives in different decision-making contexts.
{"title":"Explainable AI for government: Does the type of explanation matter to the accuracy, fairness, and trustworthiness of an algorithmic decision as perceived by those who are affected?","authors":"Naomi Aoki , Tomohiko Tatsumi , Go Naruse , Kentaro Maeda","doi":"10.1016/j.giq.2024.101965","DOIUrl":"10.1016/j.giq.2024.101965","url":null,"abstract":"<div><p>Amidst concerns over biased and misguided government decisions arrived at through algorithmic treatment, it is important for members of society to be able to perceive that public authorities are making fair, accurate, and trustworthy decisions. Inspired in part by equity and procedural justice theories and by theories of attitudes towards technologies, we posited that the perception of these attributes of decisions is influenced by the type of explanation offered, which can be input-based, group-based, case-based, or counterfactual. We tested our hypotheses with two studies, each of which involved a pre-registered online survey experiment conducted in December 2022. In both studies, the subjects (<em>N</em> = 1200) were officers in high positions at stock companies registered in Japan, who were presented with a scenario consisting of an algorithmic decision made by a public authority: a ministry's decision to reject a grant application from their company (Study 1) and a tax authority's decision to select their company for an on-site tax inspection (Study 2). The studies revealed that offering the subjects some type of explanation had a positive effect on their attitude towards a decision, to various extents, although the detailed results of the two studies are not robust. These findings call for a nuanced inquiry, both in research and practice, into how to best design explanations of algorithmic decisions from societal and human-centric perspectives in different decision-making contexts.</p></div>","PeriodicalId":48258,"journal":{"name":"Government Information Quarterly","volume":"41 4","pages":"Article 101965"},"PeriodicalIF":7.8,"publicationDate":"2024-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S0740624X24000571/pdfft?md5=79e232415d01bd4e88037e2540f7bb9f&pid=1-s2.0-S0740624X24000571-main.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142229033","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"管理学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号