If you visit Swedish schools today, you will find teachers rediscovering the ancient arts of printed books, silent reading, and handwriting practice. It has become clear that students remember less when reading from bright screens than when reading the same text on real paper.
Sweden’s educational policy revision in 2022 and 2023 cannot be seen merely as an administrative or political adjustment. It represents a deeper reconsideration of the relationship between technology, the human brain, and learning. A country that was once among the leaders in digitalizing education has now turned toward strengthening the role of printed books, offline reading, and handwriting in early grades. This shift is not a rejection of technology, but a data-driven response to a fundamental question: Can widespread and unregulated digital tool usage be linked to declining reading and writing quality?
To answer this, we must consider the historical context. In the 2010s, Sweden rapidly moved toward replacing printed books with digital content through “one student–one device” initiatives. Schools were equipped with tablets and laptops, assignments moved online, and parts of assessments were conducted digitally. Policymakers assumed that faster access to information and increased visual appeal would improve learning quality. However, after several years, signs of a gap between expectation and reality began to emerge.
The PISA 2018 and 2022 results, conducted by the OECD, showed a global decline in average reading scores. The roughly 10-point drop in the OECD average was estimated to equal the loss of nearly one-third of a school year. In Sweden, although performance remained better than in some countries, a concerning increase in students below baseline reading proficiency was observed. Complementary OECD reports also showed that students who spent excessive time using digital tools recreationally tended to perform worse in reading. While this demonstrates correlation rather than causation, the strong association was enough to spark serious policy debate.
For deeper analysis, we must turn to cognitive mechanisms. Reading is a cultural skill for which the human brain did not naturally evolve; instead, existing neural networks for vision and language have been “repurposed” for reading. These networks require sustained attention, linear processing, and the construction of mental maps of text. Meta-analyses such as Delgado et al. (2018), examining over 170,000 participants, show that reading comprehension is significantly higher with printed materials compared to digital, especially for long, analytical, or time-constrained texts. This is often explained by the “digital shallowing hypothesis”: digital environments, due to their navigational and multitasking nature, encourage more superficial processing.
Eye-tracking studies show that screen reading often involves skimming. The well-known F-pattern observed in user-experience research demonstrates that online readers tend to read the first lines carefully and then move quickly and selectively through the text. In contrast, printed books encourage more linear and deeper reading patterns. Additionally, physical elements such as page thickness and paragraph location help create spatial memory anchors. These anchors are weakened in scrolling digital environments.
From the perspective of Cognitive Load Theory, screens can increase extraneous cognitive load. Even without active notifications, users remain aware of potential interruptions. This “readiness for interruption” occupies working memory capacity and makes deep concentration more difficult. Multitasking research shows that constant task-switching carries cognitive costs and can reduce information-processing quality.
In writing, there is also strong empirical evidence. Mueller and Oppenheimer (2014) showed that students who take notes on laptops record more words but perform worse on conceptual tests compared to those who write by hand. Typing allows verbatim transcription, whereas handwriting forces processing, summarization, and conceptual reconstruction. Neuroimaging studies further show that handwriting activates broader motor and language regions and builds more durable neural connections.
However, a fair analysis must acknowledge limitations. Some studies suggest the gap between print and digital reading is narrowing, especially as readers gain more digital experience. Content design quality also matters; a simple e-book with minimal distractions differs significantly from a webpage filled with links and advertisements. Thus, the issue is not simply “digital versus print,” but the cognitive design of the learning environment.
From a policy perspective, Sweden’s decision can be viewed as an attempt to restore balance. The government announced that in early grades, printed books should be the primary learning tool, and screen use should be age-appropriate. Significant funding has been allocated to provide physical textbooks. Beyond cognitive considerations, economic factors also played a role: maintaining and updating digital hardware is costly, while printed books have longer and more predictable life cycles.
A Global Approach
Sweden’s experience is not isolated. In the United States, after waves of “paperless schools,” some districts have reinvested in printed textbooks. In France, mobile phone use has been banned in primary and middle schools. In Finland, despite advanced digital infrastructure, printed books still play a central role in early education. In South Korea, concerns about screen addiction and cognitive fatigue among students have been raised. These developments suggest a global issue rooted in the biological limits of the human brain in processing information.
So, is there a link between declining literacy and digital tool use? Current evidence suggests a conditional, correlational relationship. Limited, purposeful, and well-designed technology use can enhance performance. However, widespread, unstructured use and the complete replacement of books and handwriting in early years correlate with declines in deep reading and analytical writing indicators—and cognitive mechanisms supporting this correlation are plausible. Sweden’s experience represents a form of “policy maturity,” where early digitalization enthusiasm gives way to data-driven analysis and attention to human cognitive limits.
Bottom Line
The future of education will likely be neither fully digital nor fully traditional. The real challenge is designing an ecosystem where technology strengthens foundational literacy rather than replacing it. If the foundations of deep reading and analytical writing are weakened, even the most advanced digital tools cannot compensate. Sweden’s experience shows that education, before being a technological issue, is fundamentally a matter of human cognition.
Sources
OECD. (2023). PISA 2022 Results (Volume I): The State of Learning and Equity in Education. Paris: OECD Publishing.
Delgado, P., Vargas, C., Ackerman, R., & Salmerón, L. (2018). Don’t throw away your printed books: A meta-analysis on the effects of reading media on reading comprehension. Educational Research Review, 25, 23–38.
Mueller, P. A., & Oppenheimer, D. M. (2014). The pen is mightier than the keyboard: Advantages of longhand over laptop note taking. Psychological Science, 25(6), 1159–1168.
Mangen, A., Walgermo, B. R., & Brønnick, K. (2013). Reading linear texts on paper versus computer screen: Effects on reading comprehension. International Journal of Educational Research, 58, 61–68.
OECD. (2015). Students, Computers and Learning: Making the Connection. Paris: OECD Publishing.
Katzir, T., & Paré-Blagoev, E. J. (2006). Applying cognitive neuroscience research to education: The case of literacy. Educational Psychologist, 41(1), 53–61.
Carr, N. (2010). The Shallows: What the Internet Is Doing to Our Brains. W.W. Norton & Company.
