Data for Policy 2020 Special Track

Data, analytics and digital transformation in the private sector

Chairs: Bilal Gokpinar and Steve Yoo, UCL School of Management

Data and analytics are the next frontiers in innovation, productivity, and competition for businesses. Data and digitisation affect business processes, supply chain operations, consumer, industry and market dynamics. They threaten established business models and introduce new ways for companies to create and capture value. In addition to data generated by traditional enterprise systems, businesses increasingly capture large volumes of structured and unstructured data about their customers and operations through social media sites, mobile phones, consumer devices, as well as networked sensors, smart machines and others. Similarly, firms use increasingly sophisticated analytics to transform and improve their business capabilities.

We invite submissions to explore this emerging multifaceted phenomena with a focus on the private sector. How does the nature of businesses and business operations change in the age of digitisation and data? What kind of opportunities and challenges do data-rich environments pose to small and large enterprises? How can companies create and capture value, and enhance their productivity and competitiveness through data and analytics? How can businesses get closer to various stakeholders (e.g., consumers, suppliers, governments) with data but at the same manage corresponding risks appropriately?

We seek to explore recent advances as well as challenges in the use of data and analytics in the private sector by exploring the following topics:

  • Data, digital organisation, and business processes transformation
  • Managing people and work through analytics in the digital era
  • New data and analytics driven business models
  • From data to efficient decision-making in business
  • Data, analytics and productivity improvement in large and small companies
  • Personalisation, customisation and customer analytics
  • Data driven supply chain management
  • Planning product assortments and recommendation engines
  • Product and demand forecasting with real time data
  • New forms of innovation and entrepreneurship in the digital era
  • Enterprise software and digital transformation
  • Developing data and analytics capabilities for business impact
  • Revenue and cost improvement through data and analytics initiatives
  • Challenges for adopting analytics and data driven decision making
  • Designing effective data architecture and technology infrastructure to support analytics
  • Attracting and retaining data and analytics talent

Submissions are made via Data for Policy website.

Please send any enquiries to the track chair, Bilal Gokpinar (b.gokpinar@ucl.ac.uk).

Data Quality and Development Policy

Chair: Leigh Anderson, University of Washington

Data are intended to support evidence-based decision-making, by tracking indicators, identifying drivers of development, and predicting future scenarios according to policy and investment choices. But some core level traditional administrative and survey data are particularly limited in many low-income countries, creating an incentive for decision-makers – for convenience and cost – to use increasingly sophisticated modelling tools on particularly poor quality data. This collection of papers explores the promise and challenges for development policy from the proliferation of high dimensional “big” data, advances in machine and computational methods, and how our interpretation of development might vary according to who uses these data, for what questions, and from what sources.

Some section/chapter ideas: 

Section A: Current Data Challenges

  • SDGs and the tyranny of the dashboard (portability bias and sacrificing internal validity for comparability; central and summary statistics that miss the tails; missing data and omitted indicators)
  • The politics and public accounting of country-level statistics (stakeholders, administrative data and national statistical capacity)

Section B: New Data and Methods

  • The promise and limitations of new data sources for casual inference and prediction: survey, administrative, satellite and remotely sensed, social media, web, IoT.
  • Avoiding perpetuating biases with machine learning (data measurement error, sampling bias, algorithmic bias)
  • The ethics and economics of merging and emerging data sources: privacy and precision trade-offs, sharing, cost

Section C: Sector focus

  • Human capital: what works in education and health data and policy
  • Livelihoods: data emerging from the gig economy and policies for a new workforce
  • Food systems: the particular importance and challenge of agricultural data
  • Climate: How can data and new technologies help us collectively inform policy to mitigate and adapt to increasing climate variability and stress on our natural systems?
  • Other sectors: Security, Transportation, Democracy and Governance etc.

Submissions are made via Data for Policy website.

Please send any enquiries to the track chair, Leigh Anderson (cla@uw.edu).

Data Trusts: Democratising data governance

Chairs: Sylvie Delacroix, University of Birmingham; Neil Lawrence, University of Cambridge; Jessica Montgomery, University of Birmingham

Interactions in the workplace, at home or with public services are increasingly data-enabled, mediated by digital technologies that promise to boost economic growth and enhance societal wellbeing.

While bringing benefits across sectors and areas of life, these shifting patterns of data generation and use also expose new vulnerabilities. There are already instances of seemingly innocuous data about individuals being linked in ways that generate sensitive insights; of existing forms of discrimination being replicated in the digital sphere; and of behavioural information about individuals being used to shape significant social and political debates. These novel vulnerabilities put further pressure on existing data governance systems, many of which are creaking under the pace of technology change and the shifting patterns of data use that follow.

Existing legislative frameworks already provide a forest of data rights and responsibilities: around personal data use, intellectual property, and privacy, to name a few. Current approaches to data governance, however, grant relatively limited influence to individuals seeking to negotiate the terms of data use, or rely on citizens pursuing their individual rights through time- and energy-intensive legal processes. These asymmetries are compounded by shifting patterns of data collection and use: a data environment where data collected for one purpose might be re-used and aggregated in ways that are not anticipated at the point of collection, where harms develop from a combination of algorithmically-enabled decisions that seem individually insignificant, and where data aggregation allows a small number of companies to become powerful social and economic forces, demands fresh governance responses. Recognising that the value in data is realised through its aggregation and use, such responses must enable individuals to pool data resources for analysis, while retaining individual rights and agency in decisions about data use.

Today’s governance challenge is therefore to connect conversations about data sharing to conversations about data rights, creating infrastructures that help ensure all in society benefit from data use. Data trusts offer a mechanism that can bridge the gaps between these data sharing aspirations and rights-protections concerns.

Data trusts are legal frameworks that provide independent stewardship of data use. In recent years, these have been the subject of increasing interest, both nationally and internationally. In the UK, 2017’s Government AI strategy prioritised the development of data trusts as a means of enabling data access; since 2018, policymakers in Canada have been grappling with the questions raised by a proposed civic data trust to manage data collected from smart cities developments; and in 2019 Germany’s data ethics commission recommended that funding for further development of data trusts be prioritised.

As research and policy in this area advance, different conceptualisations of data trust are emerging: for some, these trusts focus on individual empowerment, creating structures that allow individuals to work collectively to define desirable terms of data use; for others, these trusts are frameworks to give organisations confidence to share data and enable its use.

While some progress has been made in piloting such contract-based approaches to enable data sharing between organisations in some circumstances, further action is needed to create an environment in which individuals or groups are able to influence how data about

them is used, and for what purpose. Inspired by the UK’s history of using trusts and mutualisation to facilitate collective action at times of significant asymmetries in the distribution of social and economic power, this conceptualisation of trust would offer a new route to democratise data governance. Fostering such collective action requires an understanding of the full range of legal instruments that could underpin data trusts, the forms of citizen participation that data trusts should support, and the forms of technology that can make trusts work in practice.

At this stage of debate, it is timely to take stock of what has been achieved with data trusts to date, identify lessons from existing work, and create roadmaps to drive further progress. A special track on data trusts would therefore seek to:

  • Synthesise current understandings of ‘data trusts’, taking stock of lessons learned from research and development to date;
  • Identify key questions or issues where progress is needed;
  • Engage policymakers in dialogue about the use of data trusts to tackle priority policy

Topics
and questions to explore through the track would include:

  • What are the legal foundations of data trusts, and how do these relate to existing policy and legislative frameworks? What are the cross-jurisdictional aspects that need to be considered?
  • What structures or participatory mechanisms do data trusts require to enable 
individuals to exert influence over the use of their data?
  • What technical infrastructures are necessary to make data trusts work, and what role 
can technology play in such governance mechanisms?

Submissions are made via Data for Policy website.

Please send enquiries to Sylvie Delacroix (S.Delacroix@bham.ac.uk) and/or Jessica Montgomery ( j.montgomery@bham.ac.uk).

Documenting Data and Data Science: Surfacing Data Processes and Practices

Chairs: Jenny Bunn and Elizabeth Lomas , UCL

Data has been described as the “new oil” that drives society and delivers new economic value. Like oil, data requires extensive processing in order to make it fit for many purposes and this track seeks to combat the obscuring and disappearance of these processes. In recent years, new technologies have emerged that harness the power of data. Focussing on these technologies, our view of data becomes limited – it becomes seen as just the raw material that fuels the engine of algorithms and artificial intelligence. The processes and people behind it are forgotten. Recently a number of academics have sought to remind us of the people behind the data (O’Neil, 2016; Noble, 2018), but there has been less attention paid to the processes. And yet, recent attempts to build ethical and legal frameworks to govern emerging technologies, always come back to the need to document the data and its processing. For example, a recent European report on a governance framework for algorithmic accountability and transparency, starts to flesh out a “Datasheets requirement” in the form of “a semi-structured document that asks questions such as ‘Why was the dataset created?’ How was the data collected?” (Panel for the Future of Science and Technology, 2019).

This special track seeks to bring the attention firmly back to the processes behind the data and to the need for documentation of the same. Without such documentation, the governance frameworks being envisioned will never be realised. Unless we can access a trustworthy account of where data has come from and how it has been processed, we will never be able to ensure or be sure that the right data is being used in the right way for the right reasons. This track will seek submissions that surface existing documentation practices around data and data processing. Questions to be addressed include, but are not limited to, the following:

  1. What information is routinely being recorded in practice about the origins and sourcing of data and its cleaning, pre-processing and processing?
  2. What form does this documentation take?
  3. Who has access to it and where/how is it kept?
  4. Are there existing standards or regulations that dictate certain forms of documentation and what are they?
  5. How could existing documentation practices be improved or good practice shared?

Through this track, it is hoped that a conversation can be started that will allow for discussion of this important issue and for the hidden, but vital practice of data documentation to be made visible in wider debates about data for policy.

Submissions are made via Data for Policy website.

Please send enquiries to the session chair, Elizabeth Lomas (e.lomas@ucl.ac.uk).

References:

Noble, S. U. (2018), Algorithms of Oppression: How Search Engines Reinforce Racism, NYU Press, New York.

O’Neil, C. (2016), Weapons of Math Destruction: How Big Data Increases Inequality and Threatens Democracy, Crown, New York.

Panel for the Future of Science and Technology (2019), A governance framework for algorithmic accountability and transparency, European Parliamentary Research Service, Scientific Foresight Unit (STOA), PE 624.262. Available at http://www.europarl.europa.eu/thinktank/en/document.html?reference=EPRS_STU(2019)624262

Feminicide, Data and Policy: data activism, civic information ecosystems, and public policy oriented to ending lethal gender-related violence against women

Chairs: Catherine D’Ignazio, MIT, US; Silvana Fumega, ILDA, Argentina; Helena Suárez Val, feminicidiouruguay.net and University of Warwick, UK

Abstract:

Gender-related violence against women and its lethal outcome, feminicide (or femicide), are a serious problem across the world, and in particular in Latin America and the Caribbean (LAC). In this region, following activist efforts to collect and make visible data on the issue, official statistics are now required by new legislation on the issue. Criminal reforms and other public policy oriented to ending violence against women need to be accompanied by appropriate tools for data collection. In this Special Track we invite contributions focusing on data, data practices, and policy-making about or based on data around the issue of violence against women and feminicide, in particular.

Gender-related violence against women and its lethal outcome, feminicide (or femicide)[1], are a serious problem across the world, and in particular in Latin America and the Caribbean (LAC). According to data compiled by the United Nations Economic Commission for LAC (ECLAC), at least 12 women are killed daily in the region in incidents related to their gender. Following intense and persistent activism on the issue from the feminist and women’s movement, since 2010 the number of LAC countries that have criminalized feminicide in their legislation increased from just four to all countries in the region, except Haiti and Cuba (Deus and González 2018). Feminicide is now –in most LAC countries– a more visible problem that challenges societies and governments. However, this visibility has not found a correlation in the improvement of official registration systems of these incidents. Criminal reforms and other public policy oriented to ending violence against women need to be accompanied by appropriate tools for data collection (Walby et al. 2017; Walby 2019; see also Boira et al. 2015; Suárez Val 2019), yet data about violence against women is often neglected by public authorities: they are poor quality, they are difficult to obtain, they are often contested and, as a result, policy-makers, activist, and society are unable to make good use of resources and strategies to tackle this on-going challenge.

In contrast, feminist and women activists in the region have taken upon themselves to do the work that states have neglected, collecting data about cases of feminicide from news reports and other independent sources (for examples, see Madrigal et al. 2019; Chenou and Cepeda-Másmela 2019). These mapping and monitoring efforts both highlight and attempt to overcome the inadequacy of official statistics. Data activism initiatives provide a crucial accountability and transparency function, keeping the issue in the public eye, providing statistics for media and civil society, and pressing governments for structural change (see Milan 2016; Currie et al. 2016; D’Ignazio & Klein 2020). Together with civil society efforts to develop a regional standard for official data (see Fumega 2019, August 8th), these works have revealed the need to review, support, and strengthen both official and activist data practices, while making feminicide a more visible problem in Latin America.

Motivated by an emerging collaboration between MIT, ILDA, and feminicidiouruguay.net to explore issues around data, data activism, and (open) data processes and public policy-making vis-à-vis feminicide, in this Special Track at Data for Policy 2020 we invite contributions focusing on data, data practices, and policy-making about or based on data around the issue of violence against women and feminicide, in particular.

Proposed works should cover aspects related to the status of official data on violence against women and feminicide, as well as the work of activists to supply, complement, and/or contest these statistics. Questions addressed might include: How do data and data visualisation practices help make visible and orient public policy on feminicide and other forms of gender-related violence against women? What is the relationship between official statistics and data collected by activists, journalists, and other sources? What are the gaps? What are the challenges to strengthening data collection? Is it possible to have standardized structured data at the regional level? Can AI and technology aid in collecting, estimating and circulating feminicide information? What kinds of data policies have been developed around this issue? And how have public policy and other interventions been informed by data practices around feminicide? We envisage contributions will draw from disciplines such as science and technology studies (STS), statistics, HCI, feminism, sociology, critical data studies, journalism, and political science, amongst others. We welcome contributions from scholars and practitioners.

Submissions are made via Data for Policy website.

Please send enquiries to the Session Chairs through Helena Suárez Val (Helena.Suarez-val@warwick.ac.uk).

[1] While the term used in English is “femicide”, in this text we use “feminicide” in alignment with usage in Latin American activism on the issue (Fregoso and Bejarano 2010). For a discussion and analysis of the term “feminicide”, see (Luján Pinelo 2018). Outside Latin America, other terms, such as “honor crimes” (Shalhoub-Kevorkian, 2003), may also be related and/or overlapping.

Works Cited:

  • Boira, Santiago, Chaime Marcuello-Servós, Laura Otero, and Belén Sanz Barbero. 2015. “Femicidio y feminicidio: Un análisis de las aportaciones en clave iberoamericana.” (Femicide and Feminicidio: an analysis of contributions from Ibero-America). Comunitania: International Journal of Social Work and Social Sciences 10 (July): 27–46.
  • Chenou, Jean-Marie, and Carolina Cepeda-Másmela. 2019. “#NiUnaMenos: Data Activism From the Global South.” Television & New Media 20 (4): 396–411.
  • Currie, Morgan, Britt S Paris, Irene Pasquetto, and Jennifer Pierre. 2016. “The Conundrum of Police Officer-Involved Homicides: Counter-Data in Los Angeles County.” Big Data & Society (December): 1-14.
  • Deus, Alicia, and Diana González. 2018. “Análisis de Legislación Sobre Femicidio/Feminicidio En América Latina y El Caribe e Insumos Para Una Ley Modelo.” (Analysis of Femicide/Feminicide Legislation in Latin America and the Caribbean and Inputs for a Model Law). MESECVI; ONU Mujeres. http://www2.unwomen.org/-/media/field%20office%20americas/documentos/publicaciones/20 18/12/informe%20analisis%20de%20leyes%20de%20femicidio%20en%20alc%20dec%203% 202018compressed.pdf?la=es&vs=441.
  • D’Ignazio, Catherine, and Lauren F. Klein. 2019. Data Feminism . MIT Press.
Fregoso, Rosa-Linda, and Cynthia Bejarano. 2010. Terrorizing Women: Feminicide in the Americas . Durham; London: Duke University Press.
  • Fumega 2019 (August, 8th). “Standardisation of femicide data requires a complex, participatory process, but key lessons are already emerging” LSE blog post. https://blogs.lse.ac.uk/latamcaribbean/2019/08/08/standardisation-of-femicide-data-requires-a -complex-participatory-process-but-important-lessons-are-already-emerging
  • Madrigal, Sonia, Ivonne Ramírez Ramírez, María Salguero, and Helena Suárez Val. 2019. “Monitoring, Recording, and Mapping Feminicide – Experiences from Mexico and Uruguay.” In Femicide Volume XII: Living Victims of Femicide , edited by Helen Hemblade and Helena Gabriel, XII, 67–73. Vienna: Academic Council on the United Nations System (ACUNS). http://femicide-watch.org/sites/default/files/Femicide%20XII_0.pdf#page=73.
  • Milan, Stefania. 2017. “Data Activism as the New Frontier of Media Activism.” In Media Activism in the DigitalAge, 151–63.Routledge.
  • Shalhoub-Kevorkian, Nadera. 2003. “Reexamining Femicide: Breaking the Silence and Crossing ‘Scientific’ Borders.” Signs: Journal of Women in Culture and Society 28 (2): 581–608.
  • Suárez Val, Helena. 2019 [forthcoming]. “Datos Discordantes. Información Pública Sobre Femicidio En Uruguay.” (Discordant Data. Public Information about Femicide in Uruguay). Mundos Plurales.
  • Walby, Sylvia. 2019. “Towards Zero Violence: Putting Gender into a Theory of Violence & Society.” Seminar, Department of Development Studies, SOAS University of London, March 5. https://www.youtube.com/watch?v=boGXqjMFf-8.
  • Walby, Sylvia, Jude Towers, Susie Balderston, Consuelo Corradi, Brian Francis, Markku Heiskanen, Karin Helweg-Larsen, et al. 2017. The Concept and Measurement of Violence Against Women and Men . Policy Press.

[1] While the term used in English is “femicide”, in this text we use “feminicide” in alignment with usage in Latin American activism on the issue (Fregoso and Bejarano 2010). For a discussion and analysis of the term “feminicide”, see (Luján Pinelo 2018). Outside Latin America, other terms, such as “honor crimes” (Shalhoub-Kevorkian, 2003), may also be related and/or overlapping.

‘For good measure’: The challenges of quantifying complex problems for policymaking

Chair: Sarah Giest, Leiden University

At the core of this track is the idea that even though the volume of data has increased in recent years, the quality of the data in combination with traditional administrative structures limits government’s ability to create transparent, evidence-based, accessible and responsive policies. Simply put, having a lot of data does not necessarily mean that the data are representative and reliable (Desouza and Smith 2014) or that government is able to utilize them. This is especially true for complex problems, as they are dynamic due to the number of stakeholders involved, and the feedback loops among interrelated components.

The challenge of measuring complex problems for policymaking has increasingly taken hold in the literature from different angles. The data science as well as the public policy community grapple with the intersection of data and public service delivery where bias in the data can lead to unequal treatment of citizens. So far, research shows that data mining can only address problems that lend themselves to formalization (Barocas and Selbst 2016), which means complex issues, such as poverty, are hard to grasp through data science techniques. There is also the issue that the interpretation and assessment of results are too often done by data experts, not by domain experts, which raises questions about decontextualizing information when relying on data. Finally, when applying new techniques, such as machine learning, to existing administrative datasets, scientists struggle with missing data. Methodologically speaking, this problem is often resolved by using observed proxy variables that provide probabilistic prior information about the true unobserved values (Blackwell et al. 2015). These proxy variables pose an issue in that they model characteristics more prominent in the data, which inserts a bias towards underrepresented groups. In addition, understanding the data that the government does not have may be more important in understanding what is going on than knowing the data that is available (Hand 2020).

Following from these discussions, researchers have raised accountability and transparency issues for policymaking. They highlight the stark contrast between dynamic environments with many external and changing constraints and the levels of automation for which some decision-support systems are striving (Pasquale 2015). Previous studies find that in relying on technical solutions, humans have the tendency to disregard or not search for contradictory information in light of a computer-generated solution that is accepted as correct. This is identified as automation bias (Mosier and Skitka 1996; Parasuraman and Riley 1997). Authors also point towards the importance of multidisciplinary teams of researchers, practitioners, and policy-makers in order to understand public challenges in a holistic way (Lepri et al. 2018).

In this context, the track invites conceptual and empirical papers that address a variety of issues linked to measuring complex problems for policymaking. For example:

  • Measurement challenges linked to a certain policy domain;
  • Representation and bias in government data;
  • Accountability and transparency issues linked to data;
  • The role of commercial data for policymaking;
  • Government capacity to utilize data;
  • Data gaps and evidence-based policymaking.

Submissions are made via Data for Policy website.

Please send enquiries to session chair, Sarah Giest (s.n.giest@fgga.leidenuniv.nl).

References:
Desouza, Kevin C. and Kendra L. Smith. 2014. Big Data for Social Innovation. Stanford Social Innovation Review, Summer 2014. Available at: https://ssir.org/articles/entry/big_data_for_social_innovation.
Hand, David J. 2020. Dark Data: Why What You Don’t Know Matters, A practical guide to making good decisions in a world of missing data. Princeton, NJ: Princeton University Press.
Barocas S. and A. D. Selbst. 2016. Big Data’s Disparate Impact. California Law Review, Vol. 104, No. 671.
Blackwell, Matthew, Honaker, James, and Gary King. 2015. A Unified Approach to Measurement Error and Missing Data: Overview and Applications. Sociological Methods & Research, Vol. 46, No. 3, 303-341.
Pasquale, F. 2015. The Black Box Society: The secret algorithms that control money and information. Harvard University Press.
Mosier, K. L. and L. J. Skitka. 1996. Human decision makers and automated decision aids: Made for each other? In: Parasuraman, R. and M. Mouloua (Eds.), Human factors in transportation. Automation and human performance: Theory and applications. Lawrence Erlbaum Associates, Inc., 201–220.
Parasuraman, R. and V. Riley. 1997. Humans and automation: Use, misuse, disuse, abuse. Human Factors, 39(2), 230–253.
Lepri, Bruno, Staiano, Jacopo, Sangokoya, David, Letouze, Emmanuel, and Nuria Oliver. 2018. The Tyranny of Data? The Bright and Dark Sides of Data-Driven Decision-Making for Social Good. In: Cerquitelli, Tania, Quercia, Daniele, and Frank Pasquale (Eds.), Transparent Data Mining for Big and Small Data. Routledge, 3-24.

Harnessing Data and Science to Achieve the Sustainable Development Goals

Chairs: Omar Guerrero, The Alan Turing Institute and UCL; Lorena Rivero del Paso, Global Initiative for Fiscal Transparency; Daniele Guariso, University of Sussex and The Alan Turing Institute.

In its 2030 Agenda Declaration, the United Nations acknowledge the importance of understanding development as a process with numerous interacting dimensions: “The interlinkages and integrated nature of the SDGs are of crucial importance in ensuring that the purpose of the new Agenda is realized”. Thus, when addressing the inherent complexity of the Sustainable Development Goals (SDGs), policymakers (both at high-level summits and in policy reports) discuss interdependency networks, development accelerators, bottlenecks, policy coherence, etc. Such discussions could greatly benefit from quantitative insights derived from novel data on the SDGs and new methodologies stemming from data science, complexity science, and computational social science. Bridging the gap between SDG data –generated by governments, NGOs and other organisations– and the scientific community is critical to advancing nations’ progress towards the 2030 Agenda. In fact the UN Secretary General has recently pointed out this issue in the 2019 Report on the Sustainable Development Goals: “This report therefore highlights areas that can drive progress across all 17 SDGs: better use of data; and harnessing science”.

This special track aims at bringing together experts from governments, multilateral organisations, NGOs, the private sector, and the scientific community to exchange their latest advancements in generating and harnessing data to achieve the SDGs. Since the challenges of the 2030 Agenda vary between countries, subnational regions, government levels, sectors, etc., this track seeks to provide a forum where diverse experiences and solutions cross- fertilise each other. The participation of scientists and policymakers from low and middle income countries is especially welcomed.

Topics to be covered:  

As a holistic agenda, the SDGs cover a vast space of policy topics (the 17 SDGs are further subdivided into 169 targets and 232 indicators). Therefore, this special track is agnostic about the topical application. Instead, it emphasizes aspects related to the generation of data; on the relevant challenges; on context-specific stakeholder needs; and on novel quantitative methods. While not exhaustive, the list below provides some of the topics that this special track aims to cover:

  • Assessing data-coverage across the SDGs
  • Challenges in building SDG indicators from traditional data sources
  • Compatibility between ‘official’ and ‘unofficial’ SDG indicators
  • Local or sector-level solutions for building context-specific SDG data
  • Mapping SDG indicators into public policies and policy instruments
  • Understanding the role of public finance in mapping SDGs and expenditure data
  • Measuring SDG interlinkages
  • Integrated financial frameworks for the identification of SDG interlinkages
  • Applying network science to understand the SDGs
  • Exploring new data sources and methods for building SDG indicators
  • Constructing alternative measures of sustainable development
  • Methods for aggregating indicators into targets and goals
  • Understanding the governance of SDG indicators and other related data
  • Studying the global inequalities in the generation and ownership of SDG data
  • Enhancing states’ capacities to build and analyse SDG data
  • Using AI to infer missing data
  • Predicting the behaviour of SDGs through AI
  • Causal inference with SDG data
  • Methods for exploiting big and small SDG data
  • Generating open fiscal data and mapping it into the SDGs
  • Methods for measuring policy coherence for sustainable development
  • Methods for discovering development accelerators and bottlenecks
  • Assessing the feasibility of the SDGs
  • Agent-based models of sustainable development
  • Systems dynamics models of sustainable development

Submissions are made via Data for Policy website.

Please send enquiries to the session chair, Omar Guerrero (oguerrero@turing.ac.uk).

Regulating algorithmic trading in financial markets

Chair: Gabrielle Demange, Paris School of Economics (PSE), France

The main motivation for this track is the regulation of algorithmic trading in financial markets. Algorithmic trading (which includes high frequency trading) is being increasingly used on electronic platforms. The multiplication of exchange venues and fragmentation of trading following deregulation in the 80s and the huge investments in speed have made this trading profitable. The development of this new type of traders however has raised concerns about their impact on the fairness and stability of exchanges. Long-term investors would turn to exchange venues that are less transparent than platforms to avoid being detected; algorithmic trades would have fuelled some events such as the flash crash on May 6, 2010 in the U.S financial markets.

This new environment thus calls for evolutions in the regulation and organization of exchanges to improve stability and fairness. We encourage submissions on the following two topics:

  1. Stability. How to control the potential risks generated by algorithmic trading in financial markets? So far, the regulation acknowledges this risk. To illustrate, the Markets In Financial Instruments Directive (EU MIFID) state:« An investment firm that engages in algorithmic trading shall have the effective systems and risk controls suitable to the business it operates to ensure that its trading systems are resilient … and prevent the sending of erroneous orders … that may contribute to a disorderly market. »The directive points to the contribution to a disorderly market, but is silent about the tools for mitigating such contribution. Whereas a firm is fully responsible for the sending of erroneous orders, it does not control the spill-over effects that such orders might induce in a market populated with algorithmic traders. The main issue is: How an exchange could organize the control of interacting algorithms?
  2. Fairness of exchanges. Academics have recently proposed to change the design of exchanges to diminish the advantage of being very fast. Among the proposed changes, there are the introduction of batch auctions, the replacement of the current limit orders by continuous limit order functions, or the implementation of rewards to traders posting limit orders on both sides. These studies need to be developed and tested.

Submissions are made via Data for Policy website.

Please send enquiries to the session chair, Gabrielle Demange (demange@pse.ens.fr).

Google Economics: Data – Complex Models – well-informed Policy Making

Chairs: Doyne Farmer, Institute for New Economic Thinking, University of Oxford; Claire Connelly and Carla Coburger, Rebuilding Macroeconomics, National Institute of Economic and Social Research.

More than a decade since the global financial crisis, more data is being generated about the interactions of individuals, firms, governments and organisations than ever before, yet very little is being collected. The root cause of the crisis wasn’t just reckless lending, but a lack of transparency and a failure to understand the relationships between the many agents that make up the financial system and their exposure to one another. Without fine-grained data about production networks, asset holdings, supply chains, the web of obligations and contracts linking their activities, we cannot accurately understand contagion, or how future shocks could amplify and destabilise the economy, where they might originate from and how they move through the financial system. While some researchers are developing system-wide models, much of this work is hindered by a lack of system-wide data. Concerns over privacy are hindering the ability of governments to collect data necessary for making informed policy decisions, yet, despite the implementation of new data control policies like GDPR, the private sector has more scope and freedom than ever to collect vast troves of data and are under no obligation to disclose to authorities what information they hold or how it was obtained.

Analysts are caught in a catch-22 situation: they understand the economy through the data they collect, but without new sources of data, there is no imperative to ask the new questions that would provide a more detailed understanding of how it operates. But the economy is clearly more complex, with longer supply chains and indirect financial flows within and between industries and countries. Modelling cannot keep up without the data, but without new data there is no incentive for researchers to adopt new methods and new models and so we cannot keep pace with its evolution. While some progress has been made to develop models that better detect and account for the risks that led to the crisis, the computing power of most models is spent on finding equilibrium, very little on how to get there, and almost none on tracking the flows of goods and services around the globe and the interactions of vendors, distributors, manufacturers etc.

The alternative is to acknowledge the more complex socioeconomic system by recognising the many linkages between agents other than indirectly through market prices. Once we accept that agents can trade directly, learn and change strategies in response to one another, we have a complex system where the whole is not the same as the sum of the parts. We enter the new economic reality where the economy is a complex system and displays emergent properties which cannot be deduced from individual level behaviour alone. Though there have been some improvements in recent years, the data collection and classification systems are largely reflective of a pre-digital world. We say that this cannot continue and at the same time hope to understand the evolution of the economy and society.

Current economic data collection methods are not equipped for the digital economy of the 21s t century. Data collection cannot be separated from the development of economic models – the two must proceed hand-in-hand.

We envision the concept of “Google Economics”: a world where data is publicly available and searchable by type and category on any device, by anyone. In a world of big data, artificial intelligence and the Internet of Things, (concepts we will explain in detail in further chapters), we envision collecting enough data and curating it in a way that any person could point their smartphone at a building and it would tell you about all of the economic activity occurring within its walls, including:

  • The flow of goods and services moving in-and-out of the building
  • Stocks and flows of capital
  • The web of contracts linking the activities and relationships between employers, staff, vendors, 
distributors, financiers etc.
  • Demographics and occupational capabilities: wealth and poverty
  • Ownership
  • The ecology of innovation: identifying the thinkers influencing economic activity
  • Identifying the physical and environmental impacts of economic activity: how much carbon is 
being emitted, or exploring how accomplished recycling programs and activities have been

Topics:

  1. What kind of data do we need for building complex models for well-informed policy making (Economic modelling perspective)
  2. What is stopping the collection and availability of fine-grain microdata to researchers? (culture of data perspective)
  3. How can policy makers benefit from complex models? (policy perspective)

Submissions are made via Data for Policy website.

Please send enquiries to the session chair, Carla Coburger (c.coburger@niesr.ac.uk).

Data technologies and governance frameworks used for gathering, storing, managing, processing, analyzing and sharing data in the public administrations

Chairs: Francesco Mureddu, Lisbon Council; David Osimo, Lisbon Council; Stefaan Verhulst, The GovLab, NYU; Vassilios Peristeras, European Commission.

The data explosion is affecting all aspects of the society and the economy – and public administration is no exception. Data is a fundamental resource for carrying out all government activities, from regulation to service provision. And governments everywhere and at all levels are looking into the opportunities of data driven innovation, and in many cases experimenting with it. IDC estimates that central government is the fifth largest industry of the of the big data analytics market, covering about 7% of the expenditure, and fastly growing. A recent study by Deloitte (2016) identified 103 cases of big data analytics in government. In that regard, the Communication on “Data, Information and Knowledge management” calls for a more strategic use of data, information and knowledge. In this context, a data strategy (DataStrategy@EC) and a related Action Plan have been set-up in 2018, with the objective of transforming the EC in a data-driven organisation. The eight actions of the Action Plan are centered around 5 different dimensions: data, people, technology, organisation, policy. The data strategy highlights indeed that these dimensions need to mature and evolve harmonically to deliver a real transformation on how data is used in the decision-making processes. In 2019, an operational governance framework has been set up to closely follow-up the implementation and the evolution of the Action Plan. The 2016-2020 ISA2 (Interoperability solutions for public administrations, citizens and businesses) programme funded with a budget of 131 million euro, aims to support the development of digital solutions that enable public administrations, businesses and citizens in Europe to benefit from interoperable cross-border and cross-sector public services.

All these initiatives foster data-centric public administration. But where do we stand?

To understand that the European Commission has commissioned the study Data Analytics for Member States and Citizens, which provides policy Directorate Generals of the European Commission and Member States public administrations with a knowledge base and guidance on the adoption of public sector data strategies, policy modelling and simulation tools and methodologies, and data technologies fostering a data-centric public administration.

Specifically, the study covers three domains in relation to data analytics in government:

  • Data strategies, policies and governance: initiatives in the public sector both at the strategic level, such as data strategies, data strategies, data governances and data, management plans; and at organisational level, aimed to create units or departments, and to elaborate new processes and role.
  • Policy modelling and simulation: initiatives to improve policy analysis through new data sources, robust and reliable models to perform “what-if” scenarios, predictive analytics and hypothesis testing, and tools allowing policy makers to carry out scenario analysis through intuitive interfaces.
  • Data technologies: new architectures, frameworks, tools and technologies to be used by public administrations to gather, store, manage, process, get insights and share data. This domain includes the study of how data are governed as well as data collaboratives, and in particular stresses the joint analysis of governance and technologies.

In this respect, the track will build on the results of domain 3, data technologies, and in particular on the case studies, which will be the core of the track and which will be complemented with the submissions by other participants.

The case studies that will be carried out in domain three are:

  • Reproducible Analytical Pipelines
  • Fraud Analytics at Lithuanian Customs
  • The Estonian customs analytics project
  • New Zealand Social Investment Unit
  • FINDATA and KANTA

The responsible for the case studies and the researchers drafting the case study will be the first to be invited to present at the track, and other researchers will join in the

submission as they will see that interesting presentations will already be taking place.

Submissions are made via Data for Policy website.

Please send enquiries to the session chair, Francesco Mureddu (francesco.mureddu@lisboncouncil.net).

Recent experiences using GovTech to address data sharing challenges and to implement modern data access paradigms

Chairs: Harald Stieber, European Commission; Fabio Ricciato, EUROSTAT; Jacopo Grazzini, EUROSTAT.

In this special track, we invite submissions from practitioners working on GovTech solutions within governments and international organisations, as well as from academics either contributing directly to such work via dedicated projects or addressing technological or methodological issues that have been identified through such work.

In particular, we propose three themes without being exhaustive:

  • Shared infrastructures for data analytics: different government bodies may want to leverage similar data sources for similar application, i.e., they have the same use-cases and must face the same challenges. In this context it is desirable to join forces and co-develop shared infrastructures (and tools) that are then adopted by multiple organisations. We would welcome papers reporting on early experiences in (co-) development of shared infrastructures for (big) data analytics.
  • New data access paradigms: policy makers often face situations where the data collected by one or more institution(s) contain information that are important for another institution. The traditional approach in this scenario foresees that data are moved across institutions, i.e., data access takes place in the form of data sharing. This approach is problematic where source data are confidential. We would be interested in exploring alternative paradigms that allow only the desired final information (not the source data) to flow across institutions. What technologies are available today to decouple the access/use of data that should otherwise remain confidential?
  • Impact on skills: We invite papers that explicitly take on the issue of “upgrading” and maintaining an adequate portfolio of digital skills in the public sector, preferably where this is related to themes 1) and 2).

Submissions are made via Data for Policy website.

Please send enquiries to the session chairs, Harald Stieber (harald.stieber@ec.europa.eu), Fabio Ricciato (Fabio.RICCIATO@ec.europa.eu), and Jacopo Grazzini (Jacopo.GRAZZINI@ec.europa.eu).

Data Literacy for Policy

Chairs: Annemijn van Gorp and Chantal Brakus, Department of Economic Affairs & Climate Policy, Netherlands

The private sector is rapidly becoming more data driven. Investments in data warehouses, data management processes and systems, the development of metadata standards, and the building of agile data science teams are increasingly commonplace in order to increase businesses’ efficiency, decrease costs and enhance the customer experience. In governments, the benefits of data are also increasingly recognized: the rise of data programs in CIO Offices, or the establishment of CDO Offices, are evidence of this trend. Nevertheless, the use of data by policy makers, in various parts of the policy making cycle, is lagging behind.

Across both local and national government, we observe departments’ struggling to become more data driven. Not only do they perhaps need to adhere to higher standards when it comes to ethics, open data, and data governance processes; they particularly face difficulty in creating a data culture, and in organizing their people to take up the challenges of the new data era and to use data for the complex societal challenges government face.

Becoming a data driven government, means a new way of working for policy makers, through the introduction and positioning of new roles in the organization such a data scientists, and collaborating in (new) multi-disciplinary teams, among others. Previous research has indeed discussed that policy makers’ recognition of what big data and data science can do for them remains largely unrecognized (Lutes, 2015). This makes it difficult to convince policy makers to use data. In addition, the gap between data scientists and policy makers tends to be large: policy makers themselves largely lack technical skills, and policy makers and data scientists tend to “speak different languages” (Giest, 2017, Thompson et al., 2015; Uzochukwu et al., 2016).

To promote greater use of data, or in other words, to further stimulate ‘evidence based policy’, we need to understand how governments create policy narratives, and show policy makers how evidence, or data driven policy making, helps build their case.

We propose to focus this track on how can we develop data literacy for policy making purposes. Data literacy is generally referred to as “the ability to read, work with, analyse, and argue with data.” As such it focuses on the competencies involved in working with data. Data literacy is viewed by large tech consulting firms like Gartner as a critical skill both in business and life, and as “a starting point for any data driven journey”: without data literacy, value cannot be derived from data (Gartner, 2019).

We propose to focus this track on empirical studies that provide more insight into the development of data literacy in government. Topics in this special track include, but are not limited to:

  • Agile working in disciplinary teams
  • Enacting cultural change across policy maker – management – and board levels
  • Bridging the policy & technical divide
  • Approaches to developing knowledge & skills and building competencies
  • Position of the data scientist in the organization: new organizational structures
  • Creating policy narratives: impact of data in the policy cycle
  • Data-algorithm-policy interactions

Submissions are made via Data for Policy website.

Please send enquiries to the session chairs, Annemijn van Gorp (a.f.vangorp@minezk.nl) and Chantal Brakus (c.g.brakus@minezk.nl).

Data Governance in the Public Interest

Chair: Alison Powell, LSE and the Ada Lovelace Institute

This special track focuses on issues of data access and data sharing in public life, and looks at questions of governance, infrastructural constrains, and methods to establish data stewardship that embeds ethical values or supports struggles for justice. It welcomes submissions from a broad range of disciplines, within academia, civil society, public sector or tech industry.

Addressing the problems of data under-use and data misuse is key to unleashing the social value of data. Public institutions and social actors make the best of data when it is accessed or shared across different domains. Mechanisms for data sharing require robust practices of stewardship in order to ensure that political, socio-economic and civil rights are respected, guard off possible future shortcomings and guarantee that the benefits of the services that emerge out of data sharing will be equitably distributed among people and communities. Our track starts from the consideration that the existing field of play makes the development of these mechanisms difficult, if not impossible, due to economic imbalances that empower specific actors. In the digital economy, data is seen as having financial value to extract and, to this effect, data infrastructures are set up to maximise their extractive capacity. If public and civil society actors want to unleash the social value of data, it is necessary to rethink not only how data could be governed, but also interrogate the existing structures.

Can data governance counter power imbalances? Is data governance the right framework to think through this set of concerns? What other frameworks should we advance?

There are three high level strands to frame these considerations, and we would like to encourage submissions under any or an intersection thereof:

  • Questions of public sector-specific data access and governance
    In the public sector, there are growing efforts towards data unification – within central and local government. This often involves procurement of third-party software and services. There are many consequences to these practices in terms of individuals and communities’ relations with public services as well as in terms of the interactions between policymaking and engineering. This strand encourages reflections, by way of case study or theory, on these practices of data sharing and unification, which are currently developing away from public scrutiny.This track can address issues related to problem-framing (i.e. how an issue of public interest is translated into a data-problem and why), procurement and how the power of procurement can be used to shape the market of data analytic systems, public engagement, uneven distribution of expertise among different government departments and between public and private sectors, training needs etc.
  • Questions of data access and sharing that cross public and private sectors
    Data from public life offers unique opportunities and insights that are often of interest to the private sector. This strand encourages consideration of the processes, power (im)balances, and consequences of data flows between public and private sectors.An example of a theme of interest here would be public-private data sharing agreements. While headline-cases of data sharing agreements—such as Toronto Sidewalk Lab or, more recently in the UK, the partnerships between Royal Free Hospital and DeepMind—are bound to raise concerns, other types of agreements occupy a grey zone, which is sometimes difficult to evaluate: are the agreements future-proofed? What are the trade-offs? Do they contribute to the building of public infrastructures? Is economic value produced and how? Is social value produced? How is it distributed? What should the parameters for data access and sharing agreements be? What are the more successful or interesting cases?
  • Questions of data access and sharing that cross public interest and not-for-profit organisations or research
    Data analytics that entail aggregating data sets and accessing them across different sources can significantly improve the work and services offered by civil society organizations or academic researchers in the public interest. This may involve data sources directly from government, voluntary data contributions or crowdsourcing. However, public interest purposes do not absolve, and in some cases may enhance, the need for consideration of ethical data access and governance, as well as appropriate data use. This can sometimes be challenging for those in not-for-profit contexts who may face structural obstacle due to the lack of resources, have varying levels of data literacy or require the use of commercial services.Examples of themes that could be explored in this strand include: How can we build the data governance capacity of small civil society organizations? How can we promote a safe and responsible use of data? What are the better methods to embed ethical data practices in the day-to-day use of analytics in civil society? What does ethical data sharing for research look like? What are the roles of data governance in crowd-sourced, voluntary or open-source data-for-public-interest initiatives? What are the examples of successful and productive uses?

Submissions are made via Data for Policy website.

Please send enquiries to Silvia Mollicchi (smollicchi@adalovelaceinstitute.org)

Data Governance for Innovation for Sustainable Smart Cities: Opportunities and Challenges in Public Policy and Institutional Design

Chair: Masaru Yarime, Hong Kong University of Science and Technology, Hong Kong; UCL STEaPP, UK; and The University of Tokyo, Japan.

Smart cities are expected to play a crucial role in tackling many issues concerning sustainability, ranging from reducing air pollution and increasing energy efficiency to mitigating traffic congestion and maintaining resilience to accidents and natural disasters. Data-driven innovation, including the Internet of Things (IoT), blockchain, and artificial intelligence (AI), has significant potential to address these multifaceted, interdependent issues. Vast amounts of various kinds of data are increasingly available from a variety of sources through sophisticated equipment and devices installed in buildings, automobiles, and infrastructure across cities.

Effective collection, sharing, and use of data through cooperation and collaboration among stakeholders would be critical for facilitating innovation for smart cities. While open data access and management can contribute to creating innovation, however, there are many challenges that we need to address in promoting societal benefits. There are technical issues related to data, such as metadata tagging, quality control, cleaning and error elimination, and interoperability between various standards, which must be addressed to support data sharing. Stakeholders might have different interests and motivations and would not necessarily be willing to disclose or exchange data with each other. A balance needs to be considered between open and proprietary data.

Serious concerns are also raised about collecting, sharing, and using sensitive data, particularly personal data, in terms of safety, security, and privacy. A variety of ideas are proposed for institutional arrangements for data governance in smart cities. There are debates about who should be in charge of governing data in the public or private sector. Another possibility would be to establish a data trust as an independent institution to make decisions about who has access to data under what conditions, how that data is used and shared for what purposes, and who can benefit from it.

Data-driven innovation poses a particularly difficult challenge to policymaking. The speed of technological change is rapid and the path of its evolution would not be entirely predictable or explainable, which would produce a widening gap between technological change and institutional readiness. Also, various sectors, such as energy, housing, and transportation, which were not always interconnected, are increasingly integrated through data in smart cities as cyber-physical systems. Hence institutional arrangements for data governance need to incorporate the ability to learn from real-world use and experience and the capability to improve the performance through adaptation.

A key question is what kinds of data governance systems would be appropriate to maximize the potential of innovation while minimizing risks to individuals and communities. In-depth research is required to investigate how institutional arrangements for data governance influence the collection, management, and use of data in smart cities and what impacts would be made on facilitating data-driven innovation while addressing societal concerns.

This session aims to present and share theoretical as well as empirical research findings that examine the current situations concerning the collection, sharing, and use of data in the context of smart cities and the effects of organizational and institutional arrangements for data governance on innovative efforts in the public and private sectors.

Possible questions we would discuss in this session include:

  • What kinds of data are collected in smart cities?
  • Who owns and has access to the data?
  • For what purposes are the collected data used?
  • How are the data managed by whom?
  • What incentives are provided to encourage data sharing?
  • What kinds of platforms are established to manage the data collected?
  • What impacts are made on stimulating innovation?
  • What kinds of policies and institutional arrangements are implemented to address concerns about safety, security, and privacy?

Case studies in different countries and regions are particularly welcome to understand the mechanisms and processes that promote collaboration on data, which reflect the structure of motivations and incentives that would be specific to local contexts and conditions. Various types of data governance would be addressed, including government-led, industry-led, and public-private partnership approaches. Based on theoretical and empirical research on data governance, implications for public policy and institutional design will be discussed to facilitate data-driven innovation through open data while addressing societal concerns about safety, security, and privacy in sustainable smart cities.

Submissions are made via Data for Policy website.

Please send enquiries to the session chair, Masaru Yarime (yarime@ust.hk).

Applying data for improved service design and delivery

Chairs: Barbara Ubaldi and Benjamin Welby, OECD, France

In the age of the digital transformation, the combination of data and digital technologies provide a unique opportunity for governments to improve how they design, build and run services that meet government standards while better responding to user needs. In the recent OECD Report, The Path to Becoming a Data-Driven Public Sector (https://www.oecd-ilibrary.org/sites/059814a7-en/index.html?itemId=/content/publication/059814a7-en), we observed that governments are building their capacities to develop the key enabling conditions for the application of data to anticipate, plan, deliver and monitor public services.

These conditions, established on the basis of a good data governance, enable the public sector to start actively using data to create, or increase, public value through better public services according to three areas of activity:

1. Anticipation and planning: anticipation of possible change and forecasting of needs to enabling the design of policies and services and the planning of interventions.
2. Delivery: informing and improving the implementation of policy, responsiveness of government and provision of public services.
3. Evaluation and monitoring: measuring impact, auditing decisions and monitoring performance

These areas of activity are not isolated from one another. An understanding of how data can shape future planning, present delivery and retrospective evaluation at any point in the design, delivery and implementation lifecycle of public services can support an iterative, continuously improving approach to the effectiveness of government.

We’re inviting submissions from teams and projects from the public, private and third sector with stories to tell about how data has influenced the design and delivery of services. The stories don’t necessarily have to be celebrations, perhaps your experiences came to a dead end because of a particular insurmountable challenge. What we anticipate is that through tangible case studies we can explore some of the most important questions when it comes to making better use of data to design and deliver services.

We’re interested in stories about how data can help anticipate users’ needs or enable innovative uses of technology but also how these transformative uses have recognised the importance of respecting citizen digital rights and ensuring that data for service design and delivery doesn’t damage public trust.

What problems has data solved? What has been the role of data in assessing service or policy performance either to design something entirely new or to iterate something that’s already in place? What types of data do you need and what technical problems might you need to solve? How have quantitative and qualitative data been used in combination to trigger a response or better understand needs? Have you overcome duplicated sources of data, made use of Open Government Data or benefitted from common tools like digital identity or base data registries and interoperability platforms to access data held by another organisation or part of government?

Ultimately, we want to hear about the lessons you’ve learnt which other people could benefit from. What can private sector experiences with commercial services teach the public sector, and vice versa? What do you know now that you wish someone had told you before you got started? What are the regulatory/legal/organisational/technical barriers you’ve overcome (or been thwarted by) in pursuit of delivering value to your users? What skills and disciplines have you included in your teams to maximise the use of data for designing and delivering services? How did you go about securing permission for experimentation or securing support on an ongoing basis? Do you have any bureaucracy hacks for handling business cases or benefits realisation?

Possible topics for talks could include:
• Using data for the policy or service design process
• Predictive analytics to forecast users’ needs and design/deliver customised services
• Capturing, and using, users’ feedback data for improved service design
• Specific uses of open government data in the design and delivery of services
• Ethics and privacy principles in the use of data for design and delivery of services
• The use of data for monitoring, assessing and responding to service performance (baselines, KPIs, etc.)
• Data governance and sharing models to enable service design and delivery
• The role of digital identity, interoperability platforms and other common components for supporting digital service delivery
• Developing data governance and analytics skills for service design and delivery

Submissions are made via Data for Policy website.

Please send enquiries to the session chairs, Barbara Ubaldi (barbara.ubaldi@oecd.org) and Benjamin Welby (Benjamin.WELBY@oecd.org)

 

 

 

Follow Us


Newsletter Subscription
`