Building Trust and Facilitating Use of Data

2.1 Free of political interference

Political independence must be ensured throughout the DVC, from the identification and conceptualization of what should be measured, and the design of the data collection approaches and tools, to the last phase of the DVC of data uptake and use, ensuring that data are used appropriately to make decisions and to inform policies and change, so that all can benefit.

Political interference takes many forms, from the extreme case of actual manipulation of data released by various agencies or direct interference with how data collection tools are designed, to less visible and evident forms of interference. For instance, the imposition of certain frameworks for measurement or restrictions on topics may influence the scope and content of data that limit their usability and impact, for a politically motived reason. To avoid even the appearance of political interference, data should be released simultaneously to all interested parties [5, p. 63].

2.2 Representative of all individuals

To be trusted, data must be inclusive and represent all individuals in society, including the experiences of those who are harder to reach and often misrepresented or underrepresented in data [6]. The key is to involve citizens in the decision of what is measured and how, as there is always a risk of introducing biases and excluding parts of society in the process. Citizens and communities should have agency throughout the value chain, which is particularly important for empowering marginalized groups.

In addition, data may also disguise important differences when the data collection methods do not allow for the necessary level of granularity and coverage. Data collection and analysis must proceed at a level that produces information of value to all segments of the population. This may require increasing the sample sizes of surveys or more finely disaggregating census and administrative data.

2.3 Relevant and responsive to users’ needs

Trusted data are data that respond directly to users’ needs and that answer questions from citizens, government agencies, non-governmental organizations, and decision-makers at all levels. Therefore, it is important to engage all data users in deciding what to measure, how to measure, and how to make the data available. Feedback mechanisms and active engagement with citizens are needed to ensure a continuous cycle of improvement and relevance.

2.4 Free to access, use and re-use

Free access and use are the core principles of open data. It is important for data to be free from regulations that limit their use or reuse and for data producers to make their data discoverable and provide all necessary guidance for proper and effective use, including adequate metadata and paradata, options for downloading data in machine readable formats, and tools to facilitate data use and maximize impact. A critical aspect is also the role citizens can play in decisions on improving accessibility and usability.

The misuse of data is often cited as a reason for restricting access, regulating use, or imposing penalties on data users. There needs to be a clear distinction between regulations that prevent violation of privacy and rights and ensure transparency on data practices, and regulations that ultimately restrict access. The best corrective measure to misuse or misrepresentation of statistics is public debate and citizen engagement. In addition, the Fundamental Principles of Official Statistics [7] offer important guidance on how the official statistical agencies can comment on erroneous interpretations and correct misuse of statistics. In addition, data protection regulations are needed in the context of privately held data.

2.5 Used properly and effectively

Trusted data are also data that have value and are used to their full potential by all target groups. Data may be misconstrued or misused deliberately or inadvertently. Statistical agencies have a responsibility to identify and correct misstatements [7]. They can also demonstrate the appropriate and effective use of statistics and provide guidance to others. Efforts to increase data literacy are important for increasing understanding and encouraging the use of statistics. Tools to facilitate data visualization and analysis can go a long way to maximize data use. Everyone must gain confidence to engage with and use data. Wide-spread data confidence is a building block of a fair data future.

3.1 Modernization of data governance practices

Efforts to modernize data governance have focused on: developing a better understanding of data stewardship functions and roles within the broader data ecosystem; increasing investments to improve data security and privacy practices; promoting responsible AI and ethical data practices; and building data partnerships to foster sharing and learning and bridging different data communities from both the public and the private sector.

3.2 Increasing demand for citizens’agency in data

There is wider recognition of the central role of citizens in ensuring data are trusted, inclusive and relevant to policy needs and decision making at all levels, and increasing demand for citizens’data agency, control over personal data, and intersectional data practices. The emergence of innovative approaches to engaging citizens throughout the data value chain, initiated by various parts of the data and statistical system including national statistical offices, has increased the scope of citizen engagement. These range from designing and implementing respondent-centric data collection practices to placing citizens at the center of data production, with citizens playing an active role in designing and implementing data collection initiatives. Engaging people and educating them on how to control the data flow, making them aware and giving them agency, is increasingly seen as critical to avoid attacks on democratic values and to promote trust.

A recent initiative on citizen data was endorsed by the UN Statistical Commission [16] to develop a framework that will encourage responsible production and curation of citizen data and enable CSOs, communities and citizens in countries to use the data they produce more broadly and more effectively. In that context, citizen-generated data are seen as an important tool to empower citizens to be agents of change, while enhancing data availability, timeliness, quality, openness, and inclusiveness.

Under the new initiative a framework is being developed that defines possible types of citizen data and offers a common understanding of relevant concepts and definitions as a foundation for citizen data [17]. The framework will encourage responsible production and curation of citizen data, support citizens to develop a sense of agency and ensure representation and increase the relevance and responsiveness of the NSO as the steward of the national data ecosystem by fostering an enabling environment to citizens, communities and civil society organizations as data producers.

The range of citizens’contributions to data is very wide, encompassing all stages of the data value chain and taking different forms. In developing the conceptual framework, the focus has been on data produced by citizens, communities, and civil society organizations, with the aim of informing decision making, responding to specific needs of the community, challenging existing data from the national statistical system when not reflecting the realities of a particular group or community, and filling data gaps. Other forms of citizens’contributions to regular data collection processes conducted by national statistical offices, such as censuses or surveys, in terms of engagements to increase the response rate or for other purposes, have not been included in the framework. Similarly, citizens’contributions to the analysis, dissemination and communication of data produced by government agencies, where citizens do not play a role in the production, were also excluded from what is currently covered by the framework. However, they are also important advancements in those areas, including as it relates to taking a user-centric approach in traditional data collection programmes, such as surveys, and involving citizens in efforts by National Statistical Offices to make data more relevant and accessible to users.

3.3 Increasing data transparency

Data transparency makes visible each step of the data value chain. The core requirements for data transparency are summarized by the Open Definition [18]. This requires making data freely available to the public and publishing the data, metadata, and paradata in readily accessible formats without restrictions on who can use them or how they can be used. By removing barriers to access and use, open data facilitates data integration and linkage to policy, increasing the impact and value of data.

The origins of the movement for making data open can be traced back to 1995 when the US National Research Council called for making environmental data available to the public to allow researchers to use data that transcended borders [19]. There are strong arguments for making data open, both political and ethical. For example, people have a right to see how their government is working and to use that information to demand accountability. There are also significant benefits, both economic and social, including allowing for more rapid uptake of innovative solutions, and to save lives when data are open and accessible in crises and emergencies.

The open data movement for official statistics achieved its first prominent international success in 2010 with the open release of the World Bank’s World Development Indicators [20]. However, its antecedents can be found in Principle 1 of the Fundamental Principles of Official Statistics, first adopted in 1994.11 The International Open Data Charter, adopted at the Open Government Partnership Global Summit in 2015, extends the concept of open data by incorporating principles of transparency based on comparable and interoperable data; the use of data for improved governance and citizen engagement; and inclusive development and innovation [21].

As the concept of open data has evolved, it has been an amazing enabler for moving data up the data value chain towards greater value and impact. During the COVID-19 pandemic we saw the power of agencies with open data capacity to stand out in helping with data to manage the pandemics. The same will be true of the increasing demand for climate change data and gender equality and other sectors that are still suffering from serious data gaps. Overall, open data has proved to be fundamental to building robust and sustainable data systems, promoting ethical data practices, and maintaining trust in data.

In 2019 the UN Statistical Commission established a working group on open data composed of NSOs and other partners such as Open Data Watch [22]. The working group has addressed issues of data interoperability, and implementation of open data practices and offered guidance to NSOs for the assessment and practical application of open data in the context of official statistics.

There is evidence of increasing accessibility and openness of official statistics. The Open Data Inventory (ODIN) has found slow improvement over time in the availability and openness of official statistics. In 2023 the average coverage score, measuring the availability of time-series data in 22 statistical categories, improved from 43.0 (out of 100) in 2016 to 50.6 in 2022. (See Fig. 2) Over the same period, openness scores improved from 40.3 to 52.2 [23]. For many countries, large improvements are possible by publishing more disaggregated data with longer time series and adopting fully open terms of use, but much work remains to be done.

3.4 Monitoring data transparency and openness to maintain trust

There is an increasing interest in monitoring the statistical and data capacity of countries along the data value chain from production to use and links to policy. The agencies and organizations involved in these monitoring exercises play a crucial role in promoting transparency, accountability, and building trust. They create incentives for improved governance by collecting information on the availability of government data, including official statistics, and their adherence to legal and professional standards. The results are typically summarized as indexes that provide comparable ratings across countries.

One of the earliest examples of a monitoring program was the International Monetary Fund’s Special Data Dissemination Standard and General Data Dissemination System [5], both of which were primarily concerned with the production standards and publication schedules of economic and financial data as reported by IMF member states. Summaries of the results are available through the Data Dissemination Standards Bulletin Board [24]. The self-reported results are subject to review by the IMF.

The Open Data Inventory (ODIN) produced by Open Data Watch has measured the coverage and openness of official statistics over 195 countries since 2016. ODIN is often described as a measure of the “revealed capacity” of national statistical systems because it measures what countries choose to produce with the resources available to them. ODIN assessments are made at arm’s length, based on what countries publish on official websites. Data are collected biennially by trained assessors. Subscores measuring data coverage and openness are available for 22 topical categories of statistics.

The Statistical Capacity Index (SCI) produced by the World Bank from 2004 to 2020 and now superseded by the Statistical Performance Indicators (SPI) provides a general assessment of the capacity of official statistical systems [25]. The SPI compiles information from online sources and other assessment programs, including the Open Data Index. Results are presented for five topical “pillars”: Data use; Data services; Data products; Data sources; Data infrastructure. Information to fully specify some pillars are still under development.

Based on their broad country coverage and focus on the capacity of national statistical systems, the coverage subscore of the Open Data Inventory and the SPI pillars for data sources and standards and methods were adopted as indicators for SDG 17.18.1, measuring the capacity of countries to monitor progress toward the Sustainable Development Goals [26].

Other monitoring programs look across a broader array of government information. The Global Data Barometer (GBD), the successor to the Open Data Barometer [27], is an example. The GDB is a multidimensional measure structured in four “pillars”: Governance; Capability; Data availability; Use and impact. GB also includes thematic modules of company information; land; political integrity; public finance; and public procurement. Data for the GDB are collected through a network of independent experts, complemented by government surveys and secondary data. The GDB currently includes 109 countries.

Many independent assessments of the functioning of governments include information about the performance of statistical systems. Two high-profile measures are the Mo Ibrahim Index of African Governance [28] and the Worldwide Governance Indicators (WGI) produced by the World Bank [29]. Although neither has the statistical system as its principal focus, they include inputs that reflect the production and use of official statistics that, in turn, affect their overall ratings.

More than just “league tables” conferring bragging rights on winners, objective assessments of countries’capacity to produce and use open data can be used to promote a dialogue with national government agencies and other partners on the importance of open data and on guidance on how to make their data open. They help to Identify opportunities for further improvement; raise awareness on capacity needs; and inform advocacy to increase the necessary funding.