From Averages to Agency: Rethinking Official Statistics for the 21st Century
By Stefaan Verhulst, Roeland Beerten and Johannes Jutting
Declining survey responses, politically motivated dismissals, and accusations of “rigged” numbers point to a dangerous spiral where official statistics — the bedrock of evidence-based policy — becomes just another casualty of distrust in government. In the below, we suggest a different path: moving beyond averages and aggregates toward more citizen-centric statistics that reflect lived realities, invite participation, and help rebuild the fragile trust between governments and the governed.
“What gets measured gets managed,” the adage goes. But what if what gets measured fails to reflect how people actually live, how they feel, and perhaps more importantly, what they care about? For too long, statistical agencies, the bedrock of evidence-based policymaking, have privileged averages over outliers, aggregates over anomalies, and the macro over the personal–in short, facts over feelings. The result? A statistical lens that often overlooks lived realities and held perceptions.
The strong emphasis on averages, national-level perspectives, and technocratic indicators always carried certain risks. In recent years the phrase “You can’t eat GDP” has popped up with increasing frequency: neatly constructed technical indicators often clash with lived reality, as citizens discovered during the post-COVID years of persistently high inflation for basic goods. Policies that failed to address citizen concerns have fueled discontent and anger in significant parts of the population, paving the way for a surge of populist and anti-democratic parties in both rich and poor countries. In today’s era of polycrisis, there is a growing imperative for reimagined policy processes that innovates and regains citizen trust. For that, we need to reinvent what and how we collect, interpret, use and communicate the evidence base for policies. In short, we need more trustworthy statistical foundations.
The challenge, it is important to emphasize, isn’t merely technical. It is epistemological and democratic. We face a potential crisis of inclusion and accountability, in which the question is not only how to measure, but also who gets to decide what counts as knowledge. If statistics remain too narrowly focused on averages and aggregates, they risk alienating the very citizens they are meant to serve. The legitimacy of official statistics will increasingly depend on their ability to reflect lived realities, incorporate diverse perspectives, and communicate findings in ways that resonate with public experience. In what follows, we therefore argue that, if official statistics are to remain legitimate, and trusted, they must evolve to include lived experiences — an approach that we call citizen-centric statistics.
The Limitation of Statistics Today
Traditionally, statistical agencies have operated with a mandate to produce descriptive insights that average out statistics across populations and countries — GDP, unemployment rates, inflation indices, demographic breakdowns. These metrics are vital as they provide a fundamental baseline and picture of reality. But they also have their limitations.
The key problem is that such approaches often flatten human experience into one-dimensional summaries–and this in turn has downstream effects on policy. Consider, for instance, the use of mean household income as a proxy for economic well-being. While useful in some cases, it also obscures wealth inequality, the frequent precarity of work, or the informal economy. Averages may describe the “middle,” but they rarely capture the volatility and vulnerability many experience daily.
Such shortcomings are not of course entirely unrecognized. As Stiglitz, Sen, and Fitoussi argued in “Mismeasuring Our Lives,” statistics must move beyond economic production to include well-being and sustainability. More generally, economists and others have long recognized that GDP is a simplistic metric that fails to capture the full complexity of human existence. In response, agencies like the UNDP have published the annual Human Development Report, which takes into account a far wider range of metrics.
In recent years there has also been a recognition in the world of official statistics there should be a stronger focus on distributions, rather than the average measures of social and economic wellbeing. For example, in the recently updated version of the United Nations framework for national accounts, new measures on distributional national accounts (DNA) are incorporated in the standard measuring framework. They show the distribution of characteristics such as income, consumption and savings across the full range of households. Another example is the recognition by the European Central Bank that household wealth should be measured and calculated in a distributional framework, the distributional wealth accounts (DWA). While distributional accounts can help address blind spots in traditional averages, they often remain expert-driven adjustments applied in a top-down manner. They add nuance, but they still don’t fully capture the complexity of lived reality. To bridge that gap, statistical systems must not only refine their metrics but also expand their methods. In recent times, some statistical offices have developed new ways to interact with the broader public, bringing them deeper into the statistical process. For example, the UK Office for National Statistics has developed a “personal inflation calculator” which takes into account individual spending patterns into account in the calculation of the actual inflation experienced by individual households.
Although this work is currently developed by official statistics actors in an experimental status, it is a welcome development that recognises the need for both official statistics and for policy making to take into account different experiences and inequalities in a number of key economic and social indicators. Yet experiments in tailoring statistics to personal experience — and to individual citizens — should be seen not as marginal add-ons, but as stepping-stones toward a deeper rethinking of the statistical process itself. The next step is to reimagine how citizens might engage with the design, collection, and interpretation of statistics — in short, to envision what citizen-centric statistics might look like.
Toward Citizen-Centric Statistics
What would a statistical agency look like if it treated the public not just as data points, but as partners? What would a more human–and humane–version of statistics look like? The concept of citizen-centric statistics, increasingly gaining traction in both development and policy circles, is relevant here.
A number of initiatives undertaken by the GovLab have highlighted the legitimacy and improved insights that can emerge when those affected by policy help shape the evidence that informs it. This includes work on developing more participatory forms of data governance, such as establishing a social licenses for data reuse (mechanisms designed to gauge people’s expectations and preferences around when and how their data can be re-used). During Covid, for instance, “data assemblies” were convened in New York, where residents debated conditions for the responsible re-use of cell phone mobility data. These deliberations didn’t just create guidelines — they surfaced insights about trust, equity, and context that raw data alone could not provide. The GovLab’s 100 Questions initiative provides another template, showing how regional and national-level policymaking can engage directly with citizens to help formulate the questions they believe should be answered and prioritize public resources.
Such efforts offer foundations and potential lessons for similar efforts within statistical systems. Another interesting example is offered by the Partnership in Statistics for Development in the 21st Century (PARIS21), which in 2001 launched an initiative to support the design and funding of pilot activities in low- and middle-income countries to improve trust in official statistics and NSSs. It anchored the discussions around developing trustworthy NSOs, and their role in building and sustaining trust in the broader information landscape. These multi-stakeholder discussions resulted in recognition that fragmentation of the new information landscape provides a unique opportunity for NSSs to leverage their role as stewards of a trusted data ecosystem.
The initiative had particular success in the island-nation of Vanuatu, where it helped mobilize collaboration between the local NSO and policymakers and politicians, for example by developing specific training sessions for parliamentarians. These trainings emphasized how to interpret and use statistical evidence in legislative decision-making, which in turn created stronger demand for official data. Vanuatu also piloted participatory approaches to data collection, where local communities were consulted on priority indicators.
Further notable examples can be found in the techniques used by organizations such as UNDP’s Accelerator Labs and Nesta’s Collective Intelligence Design Studio to gather stories, observations, and community signals that reveal blind spots in official statistics. Likewise, the UK Office for National Statistics (ONS) has explored citizen juries and user engagement panels to design indicators that resonate with public values. In Colombia, DANE worked with communities to revise its multidimensional poverty index based on local priorities. The “SenseMaker” approach, used by researchers in Kenya and the Philippines to collect micro-narratives at scale, is another good example. Instead of asking people to respond to predefined categories, researchers using this approach allow citizens to share their own stories and tag them using their own frameworks — resulting in what some have termed a “folk analytics” that captures individual and community experience. Indeed, Kenya has been a front-runner when it comes to integrating citizens-generated data (CGD) into official statistics: in 2022, education data from a civil-society organization was used for the first time in official reporting from the national statistical agency. Based on this work PARIS21 has developed guidelines on how to reuse citizen-generated data for official reporting by introducing a a Quality Framework for National Statistical Offices.
Taken together, these examples suggest real momentum behind a more citizen-centric approach to statistics. They demonstrate how involving citizens can increase trust, uncover blind spots, and enrich the policymaking evidence base. Yet this approach is not without risks and challenges. We explore some of these in the next section.
Risks and Challenges
The above suggests some of the tremendous potential of a more citizen-centric and participatory approach to official statistics. Yet in order for this vision to manifest, several challenges and risks must be overcome. Unlike traditional statistical systems, which operate under established protocols and institutional safeguards, citizen-led efforts are often more fluid, fragmented, and, partly as a result, also more contested. Greater openness and flexibility can foster innovation, but it also creates new vulnerabilities. To preserve the credibility of official statistics, these risks must be addressed directly. Below we highlight some of the most pressing challenges — along with some possible remedies or mitigating measures.
Representation and elite capture. Who exactly are the “citizen groups” that should be involved in citizen-centric statistics? In some cases, they may be broad-based and representative community organizations, but in others they may act more like lobby groups advancing narrow interests. This risk of elite capture persists across public, non-profit, and private sectors, threatening the legitimacy of participatory efforts and trust in both statistics and the resulting policies.
Quality and reliability. Data used for policymaking must adhere to clear quality standards. Even with the best intentions, citizen- or NGO-generated datasets may lack the methodological rigor of official statistics. Without careful protocols for quality and verification, citizen-centric data therefore risks being dismissed or, worse, misapplied.
Scalability and sustainability. Many promising initiatives — such as those listed above — remain small and localized. The challenge is how to scale them while preserving the contextual richness that makes them meaningful. This requires not just resources but also institutional buy-in, clear governance arrangements, and long-term funding mechanisms.
Governance and privacy. Statistical systems often contain and process highly personal or sensitive information. Without strong safeguards, there is a risk of misuse, breaches of confidentiality, or reinforcement of surveillance practices. Clear governance frameworks are therefore necessary to ensure that citizen engagement does not inadvertently erode trust.
Trust and legitimacy. Even when data quality is high, policymakers and the public may question its validity if the provenance is unclear, the methodology opaque, or simply if there is skepticism regarding the value of citizen participation. Building legitimacy requires transparent communication about how citizen-generated data is collected, vetted, and integrated into official systems — along, of course, with robust technical and policy safeguards to ensure quality and reliability.
These risks do not negate the value of, nor the need for, citizen-generated and participatory data. Rather, they underscore the need for careful design, strong safeguards, and clear governance. If addressed thoughtfully, the challenges can be transformed into opportunities to build more resilient, inclusive, and trusted statistical systems.
The Road Ahead: From Vision to Reality
Participatory statistics have always been important, but arguably never more so than at this critical moment. At a time when trust in institutions is fragile, policy problems are increasingly complex, and new data streams proliferate faster than governance frameworks can adapt, the need for citizen-centric statistics has become urgent. Participation is not simply a democratic add-on but a trust-building strategy: by opening statistical processes to dialogue and inclusion, agencies can strengthen their legitimacy. It is also a path to greater policy innovation and relevance. Lived experience provides the granular, contextual and place-based insights necessary to confront challenges such as climate adaptation, inequality, and digital exclusion. As sensors, social media, and other emerging sources flood the information ecosystem, citizen engagement helps define what should be measured and why, guiding policymakers through an age of overwhelming data abundance, and perhaps helping solve some of the most thorny problems of our era.
The vision of citizen-centric data begins from the recognition that, in a world saturated with data and information overload, the real scarcities we face are of meaning, agency and trust. Statistical agencies are often part of the problem because they are insufficiently focused on their clients, i.e., citizens. But if agencies and governments embrace the potential of more participatory approaches, they can be far more than mere producers of charts and spreadsheets. They can become facilitators of collective intelligence, architects of inclusion, and stewards of a more empathetic and effective public sector. This is the vision of citizen-centric statistics.
To achieve this vision, we need to cultivate new forms and mechanisms of governance, create decision accelerators and deliberative spaces for communities to define what matters and why. We must invest in data stewards and expand their roles to include connecting communities, technologists, decision-makers, and other stakeholders, and embed feedback loops to ensure people see how their input can shape policy and affect their lives.
This vision is not overly ambitious or unrealistic. The above examples–and the emergence of increasingly sophisticated methods for gathering different types of data–suggest the possibility of real change. The challenge is to ensure that data is not just about people, but with and for people. This shift from extractive to participatory models will require rethinking how statistical legitimacy is earned — not through neutrality alone, but through engagement, inclusion, and co-production.
To paraphrase Amartya Sen, the purpose of measurement is not to freeze society in numbers, but to expand the freedoms and capabilities of people. At its best, participatory statistics seeks to ensure our data reflects that ambition — and, more generally, that statistics remain not just credible and accurate, but also representative, in the service of democracy.
Stefaan Verhulst is the Co-Founder of the GovLab (NYC) and the DataTank (Brussels); Roeland Beerten is Chief Statistician, National Bank of Belgium and Johannes Jutting is Executive Head of the PARIS21 Secretariat
Opinions are those of the authors alone
