3. Knowledge and power
Scientific progress and research help us make decisions and solve problems, but they can also introduce new uncertainties. Education equips people with knowledge and helps reinforce the values and skills needed to make it meaningful in their own contexts. This chapter explores these issues through five themes:
Knowledge societies: explores the expansion of knowledge production, distribution and use.
Are the data too big to fail?: looks at Artificial Intelligence and data-driven decision making.
Opening up science: highlights emerging forms of scientific production and dissemination.
Knowledge governance: illustrates the expansion of research and how governments shape and promote it.
Speaking truth to power: examines the role of expert and lay knowledge in current democracies.
These trends are linked to education and its future through a series of questions and scenarios. Potential implications of COVID-19, both immediate- and longer-term, are discussed.
Knowledge production, like science and technology, and distribution, such as information and communication technologies, are ubiquitous. On the one hand, this entails new, powerful means to inform ourselves, make decisions and solve problems, from public policy to our everyday lives. On the other hand, it means new risks and uncertainties too, such as how to deal with abundant, sometimes fake or misleading information, or how to reap the benefits of a digitalised economy given rapidly emerging new jobs and markets. Today, governing information and knowledge effectively is central to both individuals and collectives. In education, key issues to look at include the changing nature of literacy in the 21st century, how to manage and produce research in a more open innovation environment, and how to prepare people for citizenship and democracy in a global and digital world.
Historically, individuals and organisations struggled with a lack of information and knowledge. Today, we struggle to handle its abundance: endless information is now widely available and accessible through Wikipedia, Google and more. Many aspects of our lives depend on a number of technologies connecting multiple pieces of knowledge and the ‘wisdom of crowds’ to offer intuitive solutions at the point of use, from sophisticated transport infrastructure to the daily weather forecast. The finest medical advice, the best product among seemingly infinite alternatives… all these things we have at the tap of a finger. In education, this vast, distributed and readily accessible universe of knowledge triggers essential questions about access, literacy and equity. How do we support all individuals in not only accessing digital information, but knowing what to do with it when they have it?
Note: OECD average includes data for 27 OECD countries. Where the data for countries were not consistently available in the same years, figures from the closest year are used (see StatLink for full information).
Source: OECD (2021), ICT Access and Usage by Households and Individuals (database), https://stats.oecd.org/.
In 2005, about one in two individuals across the OECD used the internet regularly, and only one in three did so daily or almost every day. Of course, back then smartphones did not exist as we know them today, with their high-speed connectivity, intuitive interfaces and access to infinite apps. In 2020, about 80% of individuals used the internet daily or almost every day on average across the OECD. A closer look at digital activity shows that, in 2005, only 40% of internet users used their internet for obtaining information about goods and services. By 2020, these figures had almost doubled, up to 75% of internet users.
The unprecedented connectivity afforded by digital technologies is redefining the ways in which knowledge is produced and circulated. Whereas only an elite few produced traditional encyclopaedias or the mass media of the 20th century (newspapers, radio and TV), today’s social media and internet sites like Wikipedia rely on the masses to generate content. For example, the number of pages in all wikis grew from about 10 000 to over 250 million in 20 years.
Wikis are just one example of the ‘wisdom of crowds’ on which we increasingly rely to make decisions. Other examples include how we use other people’s reviews on the web to assess products and services or digital videos to exercise and cook healthy meals. As digitalisation deepens, however, reaping its benefits is proving far more complex than just “pointing and clicking”. The sheer range and volume of digital information demands ever-stronger abilities to select, evaluate and use it.
Today, high-quality education means fostering strong digital literacy; equipping all learners with the competences needed to search, evaluate and use information and knowledge – as well as actively creating and communicating in a variety of formats and platforms.
Source: Wikimedia (2021), Pages to Date, All Wikis, https://stats.wikimedia.org/.
The availability of knowledge does not guarantee access to it or the skills to use it. How can we support all individuals to access and use knowledge effectively? What types of education are needed to enable students, teachers and education leaders to effectively do so?
Is more knowledge always better? What (digital) skills and attitudes are needed to effectively evaluate the quality and trustworthiness of information? To detect and discard poor quality, mis- and dis-information? How can we support teachers to validate the knowledge they use in their practice?
Our social circles influence our access to knowledge – directly, as sources, and indirectly, by shaping our motivation to know, for example. Will strong digital competences be enough to ensure equitable access? Should educational institutions work more actively to strengthen (digital) social ties? If so, how?
From cave paintings to the printing press, humans have continuously expanded their capacity to record and communicate their understanding of the world. Digitalisation is the latest step in this history. With our increasing time online, the digital footprints we leave behind (on social media, smart devices or sensors) are being gathered, organised and analysed at an unprecedented scale. Artificial Intelligence (AI) systems are used to improve and, increasingly, automate decisions in fields as diverse as agriculture, health, marketing and criminal justice. Although powerful algorithms raise issues of ethics, transparency, accountability and privacy, the collaboration between humans and machines may make the difference in finding solutions to our greatest challenges. In education, digitalisation and AI hold the promise of supporting high-quality education through the personalisation of teaching and learning.
Note: Data for 2018-2022 are projections.
Source: Sumits (28 August 2015), “The history and future of internet traffic”, https://blogs.cisco.com/, and Cisco (2016; 2017; 2018) Cisco Visual Networking Index, https://www.cisco.com/.
Observations in science, accounting systems in business, and the population census in government exemplify how data have long powered knowledge- and value-creation in society. In the 21st century, our capacity to generate and collect data is rising exponentially, with each action online generating data traces that can be harvested and analysed. In 1984, global internet traffic was 15 gigabytes per month. Internet traffic surpassed the exabyte threshold in 2001, and the zettabyte threshold in 2017. The data from transactions, production and communication processes are then analysed to give birth to new products, processes, organisational methods and markets.
Yet, making sense of this plethora of data is not a given. Increasingly, AI systems help us navigate and extract patterns from data that are otherwise immense and unstructured. Data-driven predictions reduce uncertainty and inform decisions – for example, refining medical diagnoses or advertising through individual profiling. AI systems use increasingly sophisticated and automated statistical tools for their analyses. This raises concerns around the ethics and fairness of machine-based judgments – if the data on which algorithms rely are poor or biased, so too are the information and decisions arising from them.
Research on AI is on the rise: the number of AI-related research publications has been increasing steadily in the last 40 years. Since the turn of the century, the growth has exploded: from less than 100 000 publications to more than 550 000 between 2000 and 2021, with the United States, the European Union and China producing over 70% of the publications. The research identifies great opportunities, but also risks and existential questions.
Delivering on the promise of AI to augment human capacities, such as freeing up teachers’ time to personalise education, will require addressing AI-related risks effectively. Yet, if the use of machine intelligence eventually outsources the process of knowing all together – that is, assimilating information and deploying it for us – bigger questions about human agency emerge.
Source: OECD.AI (2021), “AI Publication Time Series by Country” (indicator), OECD.AI Policy Observatory Live Data, https://oecd.ai/.
Education is increasingly becoming digitalised and ‘datafied’. What does this mean for the decision-making power of local actors (e.g. school leaders, teachers) compared to data owners/brokers (e.g. digital platforms and services)? Will power relationships between public and private providers of education shift?
AI systems are increasingly being used to make decisions in education (e.g. student assessment, allocation of pupils into schools). How effective and fair have the outcomes been so far? Are we auditing these algorithms? Which situations should prioritise human intervention?
What is the role of teachers in helping students understand the advantages, risks, threats and opportunities of AI? How should this topic be integrated in teacher education and professional development?
Science is a collective, cumulative and self-correcting enterprise. Scientific findings aren’t trustworthy because they are uncontroversial, but because they are open to scrutiny and can be verified, revised and thus improved. Retraction of a publication is one example of science’s internal quality assurance capacity. Both scientific malpractice and honest mistakes do exist, and the number of retracted scientific publications has risen over the past two decades, although they remain relatively rare. But science is not just a closed shop: digital technologies increasingly broaden access to and involvement in scientific publications and data to different stakeholders. Such efforts seek to keep improving the quality, integrity and legitimacy of scientific output. Opening up science can accelerate the translation of scientific findings into new ideas and products, which may in turn generate social and economic benefits.
Source: Sharma (2021), “Team size and retracted citations reveal the patterns of retractions from 1981 to 2020”, https://doi.org/10.1007/s11192-021-04125-4.
Science’s trustworthiness relies on its methods. At the same time, from honest mistakes to plain fraud, a number of scientific malpractices exist, ranging from overselling results and data fishing to partial publication and fabrication of data. A study including over 12 000 papers indexed by Web of Science found that the number of retracted papers increased from less than 200 in the year 2000 to over 1 200 in 2020. However, scientific output has drastically increased over the same period, and a relatively small number of authors are responsible for a disproportionate amount of retractions.
Retraction of scientific publications is both proof that mistakes take place and that the research community has tools to identify and correct them. Scientific collaboration may help prevent malpractice, as no author wants his or her name associated with wrongdoings. Between 2001 and 2019, the cumulative number of journals that published at least one article disclosing its reviewers’ identities, and/or their review reports of the article, grew from 38 to over 600, indicating a move towards greater academic transparency and accountability.
Open peer review journals represent one way in which digitalisation is making science more open and accountable. Open access to academic papers and other forms of scientific communication (e.g. blogs) allows for ideas to circulate faster and at a low cost. In addition, open access to research data facilitates the reproducibility of studies, the verification of their results and the re-utilisation of data for further research and educational purposes.
Furthermore, digital technologies allow a variety of social actors (citizens, civil society groups, industries and policy makers) to engage in scientific activities, from agenda-setting and co-production of research to the dissemination of scientific information and, of course, science learning.
Source: Wolfram et al. (2020), “Open peer review: Promoting transparency in open science”, https://doi.org/10.1007/s11192-020-03488-4.
Can direct engagement with academic research strengthen learners’ scientific knowledge and identities? Could citizen science be used for role modelling, encouraging greater female engagement in STEM? Is scientific literacy a basic democratic need?
Open access, big data and AI are transforming the ways science is produced and used. What digital skills are required for data-intensive science? What changes are needed in education and training?
Educational institutions generate more and more data on their students. Should all these data be shared more openly with different stakeholders (e.g. students, parents, media)? What impacts would this have for student-teacher, teacher-parent and other relationships? For example, what are the pros and cons of parents’ ability to monitor closely their children’s daily school life?
Addressing complex challenges such as the climate crisis and global economic governance requires political will and compromise based on precise analysis and understanding. In other words, it requires reliable knowledge. Across the OECD, research and development (R&D) activities are increasing as part of the quest for greater knowledge and innovation. For instance, the number of researchers has steadily risen over the last three decades. Meanwhile, governments continue to spend large sums of money for research and innovation in public and private sectors. An emerging concern is how to leverage such investment to build resilient and sustainable economies and societies. Promoting relevant educational research and scaling up effective innovative practices is also key for enhancing education’s quality.
Note: See OECD (2021), Main Science and Technology Indicators (database) https://stats.oecd.org/, for details about the distribution of researchers across different sectors.
Source: OECD (2021), Main Science and Technology Indicators (database), https://stats.oecd.org/.
Some have claimed that knowledge is the new gold. The number of individuals working in research activities has risen steadily over the last three decades. In 2018, an average of 9 out of 1 000 full-time employees across the OECD was dedicated to the creation of new knowledge, up from less than 6/1 000 in 1995.
Business enterprise employees accounted for over 60% of this research workforce across the OECD in 2018, compared to less than 10% working for government institutions (a share that has been declining in recent years). Researchers in academia are also an important part of the research workforce, although their share varies across countries. For example, in 2018, they represented over 50% of all researchers in the United Kingdom, about 40% in Mexico, over 30% in the European Union and Turkey, and 10% in Korea.
In decentralised market economies, businesses are key actors in research and innovation. At the same time, public funding has had and continues to play a key role in shaping the extent, nature and direction of innovation. In recent years, governments have changed the ways they support business enterprise expenditure on R&D (BERD), relying increasingly on indirect support measures like tax incentives, rather than direct support tools, such as R&D grants. Across OECD countries, tax incentives represented around 56% of total government support of business R&D in 2018, compared to 36% in 2006.
Drivers of this shift include international trade and competition rules, and the generalised perception that firms, not governments, make better decisions on which projects to invest in. But these trends might be changing: in recent years, there has been a push for governments to strategically guide private innovation efforts to where they are most needed, encouraging risk-taking R&D that private actors are not always willing to take on.
Note: This chart displays figures for 37 OECD countries (see the original source for greater detail).
Source: OECD (2021), OECD Science, Technology and Innovation Outlook 2021: Times of Crisis and Opportunity, https://doi.org/10.1787/75f79015-en.
How can we promote the production of educational research that is relevant for policy and practice? How can we promote its use? What is the role for knowledge “brokers”, who work to ensure the quality of the evidence and translate research for policy makers and practitioners?
How do different actors (e.g. private or public), institutional incentives (e.g. “publish or perish”) and relationships (e.g. hierarchies, networks, trust) shape the quality of academic research? How can we best support the production of high-quality education research?
Wisdom acquired through experience is crucial to good teaching. To what extent do knowledge management tools like school evaluation and teacher appraisal consider and leverage such tacit knowledge and expertise?
Sound political decisions rely on good information. However, experts do not always agree, especially when addressing complex, rapidly changing phenomena where robust evidence is not yet available. Here, democracy plays with a trump card: tolerance and articulation of dissent allow it to grasp collective intelligence and learn. Democracies are, like science, self-correcting; the actions of those in power – and the knowledge on which they rely – are open to debate and public scrutiny. The rise of freedom of information laws and citizen deliberative processes are becoming key to the exercise of democratic citizenship. They help maintain government transparency and fairness; safeguard the right of citizens to contest decisions that may appear unsubstantiated or serving particularistic interests; and improve policy outcomes along the way. What do these trends mean for education?
Note: Each tally represents a country and its package of right to information laws. For example, Sweden first passed laws in 1766 and counts as one. In 1951, Finland followed and raised the number to two.
Source: RTI Rating (2019), Historical Data on Country RTI Rating Scores, 2019, https://www.rti-rating.org/.
Citizenship is not only something stated in a passport. Citizenship is also actively exercised, founded on civil rights and liberties such as freedom of speech and the right to vote. In a knowledge society, the ‘right to know’ emerges as a key additional pillar. Freedom of information laws – laws granting access to information held by public bodies – are not new. Sweden’s Freedom of the Press Act dates back to 1766.
Yet it was not until 1946 that the United Nations recognised freedom of information as an integral part of the fundamental right of freedom of expression. Two years later, it became part of the Universal Declaration of Human Rights. Even though freedom of information laws vary in depth and breadth internationally, the number of OECD countries recognising a ‘right to know’ grew from one country in 1950, Sweden, to 37 in 2019.
More recently, citizens have been increasingly called upon to participate directly in the policy process, examining existing evidence to discuss and agree on potential solutions. This is the case of representative deliberative processes, such as Citizens’ Assemblies, Juries and Panels, which allow small, broadly representative groups of randomly selected citizens to produce informed policy recommendations for public authorities. Increasingly common over the last four decades, representative deliberative processes help mobilise knowledge and build consensus and legitimacy, particularly around issues that are values-based, require trade-offs and demand long-term solutions. Furthermore, they strengthen citizens’ political efficacy and civic engagement.
Recent examples of representative deliberative processes include the French Citizens’ Convention on Climate (2019-2020), or the Irish Citizens’ Assembly, which aimed to address socially divisive issues such as abortion and gender equality (2016-2018 and 2020). Education can help cultivate the essential knowledge, critical thinking and communication skills and attitudes necessary to participate in deliberative politics and society.
Note: The figure includes data for OECD member countries plus the European Union. Processes that spanned over multiple years are noted by the year of their completion (except for permanent ongoing processes).
Source: OECD (2021), OECD Database of Representative Deliberative Processes and Institutions (database), https://airtable.com/.
Does more knowledge lead to more action – and if so, what kind? How should education systems confront difficult questions about combining knowledge with values? Competing interests? Should civil disobedience be taught (not necessarily promoted) in civics?
Does a classroom culture that promotes respectful dissent and debate support better learning? If so, are learners given the time and space to form and articulate their own opinions? In your education institution, is it okay for students and teachers/professors to be wrong, to make mistakes?
What is the potential of policy co-creation (such as in representative deliberative processes) in education? By including underrepresented voices, could it increase evidence quality, inclusivity and public trust?
Trends allow us to consider what current patterns might mean for the future. But what about new patterns, shocks and surprises that could emerge over the next 15 to 20 years?
Building on the OECD Scenarios for the Future of Schooling, this section encourages readers to consider how growth could connect with education to evolve in multiple ways. Two vignettes illustrate possible stories: the Reader is invited to adapt and create new ones as desired. The next page sets out some key questions for education, and a set of potential shocks and surprises that could impact education and learning in unexpected ways. The descriptions of each scenario can be found in the Introduction of this volume.
Relevant sources
Cisco (2018), “Cisco visual networking index: Forecast and methodology, 2017-2022”, Cisco, https://www.cisco.com/.
Cisco (2017), “Cisco visual networking index: Forecast and methodology, 2016-2021”, Cisco, https://www.cisco.com/.
Cisco (2016), “Cisco visual networking index: Forecast and methodology, 2015-2020”, Cisco, https://www.cisco.com/.
OECD.AI (2021), “AI Publication Time Series by Country” (indicator), AI Policy Observatory Live Data, visualisations powered by JSI using data from MAG https://oecd.ai/ (accessed 6 December 2021).
OECD (2021), ICT Access and Usage by Households and Individuals (database), https://stats.oecd.org/ (accessed on 15 July 2021).
OECD (2021), Main Science and Technology Indicators (database), https://stats.oecd.org/ (accessed on 15 July 2021).
OECD (2021), OECD Database of Representative Deliberative Processes and Institutions (database), https://airtable.com/ (accessed 3 December 2021).
OECD (2021), OECD Science, Technology and Innovation Outlook 2021: Times of Crisis and Opportunity, OECD Publishing, Paris, https://doi.org/10.1787/75f79015-en.
OECD (2020), "Building digital workforce capacity and skills for data-intensive science", OECD Science, Technology and Industry Policy Papers, No. 90, OECD Publishing, Paris, https://doi.org/10.1787/e08aa3bb-en.
OECD (2020), Innovative Citizen Participation and New Democratic Institutions: Catching the Deliberative Wave, OECD Publishing, Paris, https://doi.org/10.1787/339306da-en.
OECD (2020), The Digitalisation of Science, Technology and Innovation: Key Developments and Policies, OECD Publishing, Paris, https://doi.org/10.1787/b9e4a2c0-en.
OECD (2019), Artificial Intelligence in Society, OECD Publishing, Paris, https://doi.org/10.1787/eedfee77-en.
OECD (2019), Measuring the Digital Transformation: A Roadmap for the Future, OECD Publishing, Paris. https://doi.org/10.1787/9789264311992-en.
OECD (2015), Data-Driven Innovation: Big Data for Growth and Well-Being, OECD Publishing, Paris, https://doi.org/10.1787/9789264229358-en.
RTI Rating (2019), Historical Data on Country RTI Rating Scores, 2019, https://www.rti-rating.org/.
Rubio, F. D., and P. Baert (Eds.) (2012), The Politics of Knowledge, Routledge, London.
Sharma, K. (2021), "Team size and retracted citations reveal the patterns of retractions from 1981 to 2020”, Scientometrics, Vol. 126/10, pp.8363-8374, https://doi.org/10.1007/s11192-021-04125-4.
Sumit, A. (28 August 2015), “The history and future of internet traffic”, Cisco blogs, https://blogs.cisco.com/.
Wikimedia (2021), Pages to Date, All Wikis, https://stats.wikimedia.org/ (accessed on 6 December 2021).
Wolfram et al. (2020), “Open peer review: Promoting transparency in open science”, Scientometrics, Vol. 125, pp.1033-1051, https://doi.org/10.1007/s11192-020-03488-4.
Glossary
Algorithm: A set of calculation or operation instructions for specific tasks, especially for computers. These can be simple processes, such as multiplying two numbers, or complex ones, such as playing a compressed video file. For example, search engines use algorithms to display the results from their search index for specific queries in a particular order, using criteria like relevance.
Artificial intelligence (AI) system: A machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments.
Citizenship: One’s formal legal and political status and a sense of belonging. It entails the right and responsibility to make rights claims regarding issues that affect one’s well-being.
Data fishing: The misuse of data by performing many statistical tests and only reporting those that present significant results.
Datafication: The rendering of social and natural worlds in machine-readable digital format.
Democracy: Political system in which citizens are allowed to participate in decision making and discussions. Politicians are typically elected by the citizens in free and fair elections, and serve therefore as representatives of the people.
Digital literacy: Having the competences needed to live and work in a society where communication and access to information increasingly take place through digital technologies such as the internet and smartphones.
Digitalisation: The use of digital technologies and data as well as their interconnection, which results in new activities or changes to existing ones.
Freedom of information laws: Laws aimed at allowing citizens to have access to information and data held by public entities in order to guarantee maximum transparency and accountability of government operations, encourage the reuse of information, and generate social and economic value.
Information: A collection of data that is meaningful. Information has added context to data and it can be shared (e.g. through writing or teaching).
Knowledge: Information that has been retained with an understanding about the significance of that information. Knowledge includes information gained by experience, study, familiarity, association, awareness and/or comprehension.
Market economy: An economic system in which goods and services are made, sold, and shared and prices set by the balance of supply and demand.
Open peer review journals: Academic journals that disclose the identities of their articles’ reviewers and/or the review reports of the articles.
Partial publication: Not publishing or only partially publishing the complete datasets or research material needed to validate research through a replication study.
Research and development (R&D): Research and creative work conducted by either the private and/or the public sector to develop new goods, techniques and services, and to increase the stock of knowledge and the use of this knowledge to devise new applications.
R&D grants: Financial support, typically awarded by governments, to a person or company in order to fund research and development activities.
Representative deliberative processes: A randomly selected group of people who are broadly representative of a community spending significant time learning and collaborating through facilitated deliberation to form collective recommendations for policy maker.
Researchers: Professionals engaged in the conception or creation of new knowledge, products, processes, methods and systems, as well as in the management of the projects concerned.
Retracted scientific publications: Articles that are withdrawn from the journal in which they appeared after original publication. Retractions are issued through a decision made by the publication's editorial board.
Tax incentives: A governmental measure that is intended to encourage individuals and businesses to spend money or to save money by reducing the amount of tax that they have to pay.
Universal Declaration of Human Rights: A document adopted by the United Nations General Assembly affirming an individual’s rights, such as the right to life and freedom from slavery.
Wikipedia: A free, multilingual online encyclopaedia written and maintained by a community of volunteer contributors through a model of open collaboration.