Utilizing big data without domain knowledge impacts public health decision-making


Paper by Miao Zhang, Salman Rahman, Vishwali Mhasawade and Rumi Chunara: “…New data sources and AI methods for extracting information are increasingly abundant and relevant to decision-making across societal applications. A notable example is street view imagery, available in over 100 countries, and purported to inform built environment interventions (e.g., adding sidewalks) for community health outcomes. However, biases can arise when decision-making does not account for data robustness or relies on spurious correlations. To investigate this risk, we analyzed 2.02 million Google Street View (GSV) images alongside health, demographic, and socioeconomic data from New York City. Findings demonstrate robustness challenges; built environment characteristics inferred from GSV labels at the intracity level often do not align with ground truth. Moreover, as average individual-level behavior of physical inactivity significantly mediates the impact of built environment features by census tract, intervention on features measured by GSV would be misestimated without proper model specification and consideration of this mediation mechanism. Using a causal framework accounting for these mediators, we determined that intervening by improving 10% of samples in the two lowest tertiles of physical inactivity would lead to a 4.17 (95% CI 3.84–4.55) or 17.2 (95% CI 14.4–21.3) times greater decrease in the prevalence of obesity or diabetes, respectively, compared to the same proportional intervention on the number of crosswalks by census tract. This study highlights critical issues of robustness and model specification in using emergent data sources, showing the data may not measure what is intended, and ignoring mediators can result in biased intervention effect estimates…(More)”

Augmenting the availability of historical GDP per capita estimates through machine learning


Paper by Philipp Koch, Viktor Stojkoski, and César A. Hidalgo: “Can we use data on the biographies of historical figures to estimate the GDP per capita of countries and regions? Here, we introduce a machine learning method to estimate the GDP per capita of dozens of countries and hundreds of regions in Europe and North America for the past seven centuries starting from data on the places of birth, death, and occupations of hundreds of thousands of historical figures. We build an elastic net regression model to perform feature selection and generate out-of-sample estimates that explain 90% of the variance in known historical income levels. We use this model to generate GDP per capita estimates for countries, regions, and time periods for which these data are not available and externally validate our estimates by comparing them with four proxies of economic output: urbanization rates in the past 500 y, body height in the 18th century, well-being in 1850, and church building activity in the 14th and 15th century. Additionally, we show our estimates reproduce the well-known reversal of fortune between southwestern and northwestern Europe between 1300 and 1800 and find this is largely driven by countries and regions engaged in Atlantic trade. These findings validate the use of fine-grained biographical data as a method to augment historical GDP per capita estimates. We publish our estimates with CI together with all collected source data in a comprehensive dataset…(More)”.

Place identity: a generative AI’s perspective


Paper by Kee Moon Jang et al: “Do cities have a collective identity? The latest advancements in generative artificial intelligence (AI) models have enabled the creation of realistic representations learned from vast amounts of data. In this study, we test the potential of generative AI as the source of textual and visual information in capturing the place identity of cities assessed by filtered descriptions and images. We asked questions on the place identity of 64 global cities to two generative AI models, ChatGPT and DALL·E2. Furthermore, given the ethical concerns surrounding the trustworthiness of generative AI, we examined whether the results were consistent with real urban settings. In particular, we measured similarity between text and image outputs with Wikipedia data and images searched from Google, respectively, and compared across cases to identify how unique the generated outputs were for each city. Our results indicate that generative models have the potential to capture the salient characteristics of cities that make them distinguishable. This study is among the first attempts to explore the capabilities of generative AI in simulating the built environment in regard to place-specific meanings. It contributes to urban design and geography literature by fostering research opportunities with generative AI and discussing potential limitations for future studies…(More)”.

The future of agricultural data-sharing policy in Europe: stakeholder insights on the EU Code of Conduct


Paper by Mark Ryan, Can Atik, Kelly Rijswijk, Marc-Jeroen Bogaardt, Eva Maes & Ella Deroo: “n 2018, the EU Code of Conduct of Agricultural Data Sharing by Contractual Agreement (EUCC) was published. This voluntary initiative is considered a basis for rights and responsibilities for data sharing in the agri-food sector, with a specific farmer orientation. While the involved industry associations agreed on its content, there are limited insights into how and to what extent the EUCC has been received and implemented within the sector. In 2024, the Data Act was introduced, a horizontal legal framework that aims to enforce specific legal requirements for data sharing across sectors. Yet, it remains to be seen if it will be the ultimate solution for the agricultural sector, as some significant agricultural data access issues remain. It is thus essential to determine if the EUCC may still play a significant role to address sector-specific issues in line with the horizontal rules of the Data Act. During six workshops across Europe with 89 stakeholders, we identified how the EUCC has been (1) received by stakeholders, (2) implemented, and (3) its future use (particularly in response to the Data Act). Based on the workshop results and continued engagements with researchers and stakeholders, we conclude that the EUCC is still an important document for the agricultural sector but should be updated in response to the content of the Data Act. Hence we propose the following improvements to the EUCC: 1. Provide clear, practical examples for applying the EUCC combined with the Data Act; 2. Generate model contractual terms based on the EUCC provisions; 3. Clarify GDPR-centric concepts like anonymisation and pseudonymisation in the agricultural data-sharing setting; 4. Develop a functional enforcement and implementation framework; and 5. Play a role in increasing interoperability and trust among stakeholders…(More)”

It’s time we put agency into Behavioural Public Policy


Article by Sanchayan Banerjee et al: “Promoting agency – people’s ability to form intentions and to act on them freely – must become a primary objective for Behavioural Public Policy (BPP). Contemporary BPPs do not directly pursue this objective, which is problematic for many reasons. From an ethical perspective, goals like personal autonomy and individual freedom cannot be realised without nurturing citizens’ agency. From an efficacy standpoint, BPPs that override agency – for example, by activating automatic psychological processes – leave citizens ‘in the dark’, incapable of internalising and owning the process of behaviour change. This may contribute to non-persistent treatment effects, compensatory negative spillovers or psychological reactance and backfiring effects. In this paper, we argue agency-enhancing BPPs can alleviate these ethical and efficacy limitations to longer-lasting and meaningful behaviour change. We set out philosophical arguments to help us understand and conceptualise agency. Then, we review three alternative agency-enhancing behavioural frameworks: (1) boosts to enhance people’s competences to make better decisions; (2) debiasing to encourage people to reduce the tendency for automatic, impulsive responses; and (3) nudge+ to enable citizens to think alongside nudges and evaluate them transparently. Using a multi-dimensional framework, we highlight differences in their workings, which offer comparative insights and complementarities in their use. We discuss limitations of agency-enhancing BPPs and map out future research directions…(More)”.

The Complexities of Differential Privacy for Survey Data


Paper by Jörg Drechsler & James Bailie: “The concept of differential privacy (DP) has gained substantial attention in recent years, most notably since the U.S. Census Bureau announced the adoption of the concept for its 2020 Decennial Census. However, despite its attractive theoretical properties, implementing DP in practice remains challenging, especially when it comes to survey data. In this paper we present some results from an ongoing project funded by the U.S. Census Bureau that is exploring the possibilities and limitations of DP for survey data. Specifically, we identify five aspects that need to be considered when adopting DP in the survey context: the multi-staged nature of data production; the limited privacy amplification from complex sampling designs; the implications of survey-weighted estimates; the weighting adjustments for nonresponse and other data deficiencies, and the imputation of missing values. We summarize the project’s key findings with respect to each of these aspects and also discuss some of the challenges that still need to be addressed before DP could become the new data protection standard at statistical agencies…(More)”.

Artificial Intelligence for the Internal Democracy of Political Parties


Paper by Claudio Novelli et al: “The article argues that AI can enhance the measurement and implementation of democratic processes within political parties, known as Intra-Party Democracy (IPD). It identifies the limitations of traditional methods for measuring IPD, which often rely on formal parameters, self-reported data, and tools like surveys. Such limitations lead to partial data collection, rare updates, and significant resource demands. To address these issues, the article suggests that specific data management and Machine Learning techniques, such as natural language processing and sentiment analysis, can improve the measurement and practice of IPD…(More)”.

Toward a citizen science framework for public policy evaluation


Paper by Giovanni Esposito et al: “This study pioneers the use of citizen science in evaluating Freedom of Information laws, with a focus on Belgium, where since its 1994 enactment, Freedom of Information’s effectiveness has remained largely unexamined. Utilizing participatory methods, it engages citizens in assessing transparency policies, significantly contributing to public policy evaluation methodology. The research identifies regional differences in Freedom of Information implementation across Belgian municipalities, highlighting that larger municipalities handle requests more effectively, while administrations generally show reluctance to respond to requests from perceived knowledgeable individuals. This phenomenon reflects a broader European caution toward well-informed requesters. By integrating citizen science, this study not only advances our understanding of Freedom of Information law effectiveness in Belgium but also advocates for a more inclusive, collaborative approach to policy evaluation. It addresses the gap in researchers’ experience with citizen science, showcasing its vast potential to enhance participatory governance and policy evaluation…(More)”.

Private sector trust in data sharing: enablers in the European Union


Paper by Jaime Bernal: “Enabling private sector trust stands as a critical policy challenge for the success of the EU Data Governance Act and Data Act in promoting data sharing to address societal challenges. This paper attributes the widespread trust deficit to the unmanageable uncertainty that arises from businesses’ limited usage control to protect their interests in the face of unacceptable perceived risks. For example, a firm may hesitate to share its data with others in case it is leaked and falls into the hands of business competitors. To illustrate this impasse, competition, privacy, and reputational risks are introduced, respectively, in the context of three suboptimal approaches to data sharing: data marketplaces, data collaboratives, and data philanthropy. The paper proceeds by analyzing seven trust-enabling mechanisms comprised of technological, legal, and organizational elements to balance trust, risk, and control and assessing their capacity to operate in a fair, equitable, and transparent manner. Finally, the paper examines the regulatory context in the EU and the advantages and limitations of voluntary and mandatory data sharing, concluding that an approach that effectively balances the two should be pursued…(More)”.

Collaboration in Healthcare: Implications of Data Sharing for Secondary Use in the European Union


Paper by Fanni Kertesz: “The European healthcare sector is transforming toward patient-centred and value-based healthcare delivery. The European Health Data Space (EHDS) Regulation aims to unlock the potential of health data by establishing a single market for its primary and secondary use. This paper examines the legal challenges associated with the secondary use of health data within the EHDS and offers recommendations for improvement. Key issues include the compatibility between the EHDS and the General Data Protection Regulation (GDPR), barriers to cross-border data sharing, and intellectual property concerns. Resolving these challenges is essential for realising the full potential of health data and advancing healthcare research and innovation within the EU…(More)”.