Article content
Reading the news, posting on social networks, shopping at a mall. Different activities for various purposes, but they all have something in common – human actions, generating data. In the information society, data is the footprint of a multi-layered reality in motion. However, this huge share of knowledge we create on a daily basis needs to be described, analysed. In this way, from sociodemographic indicators to economic performance, statistics become powerful tools for business and policy-making alike.
The increasing complexity of digital life calls different actors to take a leading role in deciphering and interpreting core data. Among them, national statistical offices (hereinafter “NSOs”) notably figure as the official sources of indicators and state-of-the-art information. And Statistics Estonia, the government agency led by Mart Mägi, is aiming to become the most innovative and effective producer of reliable statistics in Europe by 2022.
Signs of this new protagonism are recommendations recently submitted to the United Nations Economic Commission for Europe on the future of statistics in a changing data ecosystem. Director General Mart Mägi, in line with OECD guidelines, makes the case for smart data management as the way forward. In our interview, he explains how integrating technology with new analytical approaches can lead to more insightful, conscious decision-making – both in the public and private sectors.
Mart Mägi, Director General of Statistics Estonia
A new data ecosystem is taking shape, from sources to indicators. How can we describe this changing context, and what challenges NSOs face as the provider of official statistics?
The greatest driver of change in the data ecosystem has been the data revolution of the last several years. Two main factors are the increasing digitisation of information and the emergence of the concept of ‘big data’. The data revolution has also resulted in a change in stakeholders’ expectations – the demand for more open data, available faster, and accessible through web portals, APIs, and sharing platforms. This has resulted in the creation of a statistical output by other organisations that are leveraging the availability of new data sources, tools and techniques to generate relevant information.
The fast-paced change in technology, including artificial intelligence (AI) and machine learning, drives the demand for access to data. Increasing data literacy skills of the data analytics and data science communities also play a role. As a provider of official statistics, we should take this into account and take our operations to a new level.
Statistics Estonia is the core agency for data collection and management in our country. What is the value offer of your organization for governments, media, and private actors?
Statistics Estonia engages in new tasks based on the recently reviewed Estonian digital strategy (Digital Agenda 2020 for Estonia). The strategy focuses on the well-functioning and secure use of smart ICT solutions to increase the quality of life, productivity in the economy, and efficiency of the public sector. Within that framework, Statistics Estonia is mentioned explicitly as a “public-sector competence centre for data governance and data science, which would support data-driven decision-making in the country”.
In the new data governance framework, our agency will be involved in effective dissemination, open data and data life-cycle management. Above all, we will be committed to active data stewardship across the administration. This entails a series of core, value-adding responsibilities including:
- Supporting high quality and optimized use of data;
- Facilitating access to data;
- Promoting expertise, skills and data literacy;
- Promoting common standards, frameworks and data policies;
- Elaborating data strategies, including sharing and collaboration aspects.
The value of such tasks is enormous, not only for providing better information for knowledge-based decision-making. Finally, we need to make sure that data is efficiently used across many domains. With this respect, setting up proper metadata management and data quality becomes of paramount importance for all parties.
Data governance and data for governance. What are the principles of a new role for NSOs, and how can this impact policy- and decision-making in society?
Through NSOs, data becomes meaningful information. With a long history of handling data, NSOs have gained skills to provide high-quality data for policy- and decision-making. However, it should be noted – due to a significant increase in data use – that this process is not self-explanatory.
Handling data requires skills for data digestion and protection, as well as analytical and visualisation skills. We believe that, soon, every organisation will establish a role for data stewards. There is a need to define what data means in systems, and how others can access that relevant information. Securing that data is safely held, and at the same time provided for rightful decision-making or e-services, could be easily quantified in millions of net worth.
Is the concept of ‘big data’ slowly falling in disuse? Some propose a shift to ‘smart data’ management instead, but what advantages would that carry?
As observed in the projects analysed, value is not necessarily about the data being ‘big’. The most common way to characterise ‘big data’ is by 3 Vs (volume, variety and velocity). Sometimes, we add two more Vs (veracity and value). Value results from taking creative angles on data, whether by combining existing (small and big) data sources in new ways, as well as tapping into unconventional sources and forging new methods and algorithms.
Value also arises from matching more traditional statistical approaches with the newest data science techniques, rather than opposing them. From that perspective, ‘smart data’ represents a semantic advocacy for a value proposition that combines and transcends both types of expertise in their current roles, bringing together the statistician and the data scientist.
On par with the increase in data sources, computing capabilities progress as well. How does the future of data analysis and dissemination look like?
The data revolution enables increased timeliness and granularity in data analysis. Eventually, this will facilitate more detailed information on socio-economic and sustainable development indicators.
With an increasing amount of data and information, finding the most timely, reliable and transparent information becomes vital. Therefore, our discussion in the future will be more about what are the most important indicators for well-being and sustainable development – and not where or what data can be found.
