Screens, Books, and the Stories We Tell: A Two-Country Autopsy of Nordic Literacy Decline

---

1. Two Countries, One Story — Or Are They?

In the spring of 2025 and into 2026, two related but distinct narratives have circulated across European educational media with unusual velocity.

The first centres on Norway. The claim, repeated across Polish, German, British and American outlets, runs approximately as follows: Norway gave iPads to every schoolchild, reading scores collapsed, and the country now faces a functional literacy crisis. The implied causal chain — device → illiteracy — is stated or insinuated with confidence that the underlying data do not support.

The second, newer narrative centres on Sweden. A May 2026 Economist feature describes Nordic schools “going unplugged,” led by Sweden’s higher-education minister Lotta Edholm and her government’s “från skärm till pärm” (from screen to binder) programme. The framing is broadly sympathetic: Sweden is correcting a mistake, investing in books, banning phones, returning to basics. The implicit message — reinforced by similar moves in Finland and Denmark — is that digital education was an error and the era of the classroom tablet is ending.

Both narratives are more complicated than they appear. But they are complicated in different ways, and treating them as the same story produces its own distortions. Norway’s story is primarily about a real trend being misdiagnosed. Sweden’s story is primarily about a legitimate policy response that overstates its own causal premises. Understanding the difference matters — because the policy consequences of misdiagnosis are different from the policy consequences of overcorrection.

---

2. Norway: A Real Decline, A Wrong Explanation

2.1 What PIRLS 2021 Actually Shows

The Progress in International Reading Literacy Study (PIRLS), conducted every five years by the IEA and the TIMSS & PIRLS International Study Center at Boston College, is the world’s most comprehensive assessment of fourth-grade reading literacy. The 2021 cycle covered 57 countries, over 400,000 students, 380,000 parents, 20,000 teachers, and 13,000 schools.

Norway’s results are genuinely mixed — but substantially different from the media characterisation:

• Norway ranked 18th out of 65 participating countries and scored above the international average.
• In PIRLS 2016, Norway had ranked among the top ten countries globally, alongside Singapore, Russia, Ireland, Finland, and Poland.
• The share of Norwegian students reading at a low or below-low level approximately doubled between 2016 and 2021.
• The 2016–2021 decline was the steepest among Nordic countries.

“18th in the world and above the international average” is a substantially different claim from “a nation of illiterates.” The collapse from top 10 to top 20 within the global upper tier is a worrying trend — but the distinction between these characterisations matters critically for diagnosis and policy. Removing the baseline context transforms a genuine concern into an apparent civilisational crisis.

2.2 The Variable That Media Coverage Consistently Omits

PIRLS 2021 was the first major international reading assessment conducted entirely during and immediately after the COVID-19 pandemic. The IEA itself noted that pandemic disruption was the dominant contextual factor across participating countries. Of 32 countries with comparable trend data, 21 showed lower scores in 2021 than in 2016.

A 2025 study published in Learning and Instruction, drawing on data from 187,386 students across 18 European countries using propensity score weighting to disentangle COVID effects from other variables, found:

“

The primary factor explaining the adjusted decline in reading literacy across European countries between 2016 and 2021 was the duration of COVID-19-related school closures. Countries with zero school closures achieved outcomes consistent with pre-pandemic trend projections.

”

— Ludewig, Strietholt & McElvany (2025), Learning and Instruction

Between 2016 and 2021, both tablet deployment and a global pandemic disrupted Norwegian schools simultaneously. The media narrative attributes the outcome to one variable while systematically ignoring the other. This is not a matter of interpretation. It is a logical error.

2.3 The Cross-National Disconfirmation

The causal claim faces a further problem: it does not survive cross-national comparison. The three highest-performing countries in PIRLS 2021 — Singapore (rank 1), Ireland (rank 2 in most analyses), Hong Kong SAR (rank 3) — all have extensive and well-documented programmes of digital device deployment. Singapore’s national EdTech strategy places a 1:1 device-to-student ratio as a central goal. Ireland distributed devices to primary students as part of its Digital Strategy for Schools.

If the causal chain were “iPad → illiteracy,” we would expect these countries to show deficits proportional to their device saturation. They lead the world. This disconfirming pattern is systematically absent from media coverage.

2.4 Norway’s Actual Causes: A Multifactorial Account

Norwegian educational researchers and the Norwegian Screen Use Committee (Skjermbrukutvalget) offer a more disaggregated account.

Changing student population composition. Between 2012 and 2022, the proportion of students from migration backgrounds in Norwegian schools rose from approximately 10% to 16%. PISA 2022 data show a reading gap of 51 PISA points between native-born Norwegian students and those with a migration background — equivalent to more than one year of schooling. Even after controlling for socioeconomic factors, a statistically significant 23-point gap remains. This compositional shift is large enough to explain a significant portion of the aggregate score change without reference to any technology.

Insufficient teacher preparation for digital pedagogy. Research on Norway’s COVID-19 educational response identified a structural weakness predating the pandemic: Norwegian teachers received minimal guidance on digital pedagogy at school and national level. The GrunnDIG project (2022–2023) found that teacher educators themselves have paradoxically limited access to digital learning materials — making it structurally impossible to prepare future teachers to deploy technology critically.

Displacement of sustained reading time. The Skjermbrukutvalget’s 2023 conclusion is carefully worded: “A preliminary conclusion from the committee is that there is much to suggest that students should read more on paper. However, the research field is falling behind and provides minimal insight into long-term consequences.” Note the epistemic register: preliminary, much to suggest, results differ. This is Norway’s own expert body — not a finding that iPads cause illiteracy, but a finding that long-form continuous reading requires sustained practice, and that if devices replace rather than supplement that practice, it diminishes.

The technology itself was not designed for pedagogical substitution. Classroom research at Norwegian primary schools found that students experienced tablet use as engaging but sometimes disconnected from their existing literacies — and called for greater integration of students’ own digital competencies into classroom practice (Bjørgen, 2021). Longitudinal observation at Norwegian lower secondary schools found device use was predominantly substitutive rather than transformative: students completed on screens the same tasks they would previously have completed on paper, without pedagogical redesign (Kure, Blikstad-Balas & Brevik, 2025). This is a pedagogy failure, not a technology failure.

---

3. Sweden: A Legitimate Response Built on a Questionable Premise

3.1 Sweden Is Structurally Different From Norway

Sweden’s case requires separate analysis for three reasons. First, its PISA trajectory is different. Second, its policy response is more substantive than any media narrative. Third, some of that response is actually defensible.

The PISA data for Sweden tell a story that predates COVID and predates mass tablet deployment:

PISA cycle	Reading score	Change
2009	497	—
2012	483	–14
2015	500	+17
2018	506	+6
2022	487	–19

Two observations are immediately visible and both challenge the “devices caused this” narrative.

The steepest decline in Swedish reading performance — 14 points between 2009 and 2012 — occurred before mass tablet deployment, which accelerated from 2013–2016. The best recent performance (2015–2018, +23 points of recovery) coincided with the peak of the device rollout era. The 2018–2022 drop is real and significant — but it happened during a pandemic, and the Ludewig et al. finding about COVID school closures as primary driver applies here as well.

Lotta Edholm’s statement that the decline is “linked to an overreliance on digital tools without sufficient evidence of their effectiveness” is a political claim dressed as an empirical one. The PISA trajectory does not support it. What it supports is a more cautious formulation: we don’t know precisely how much devices contributed, but we know that more structured reading is good, and that unstructured phone use during lessons is harmful. So we are acting on what we know. That would be defensible. The stronger causal claim is not.

3.2 Sweden’s Structural Problems That Devices Did Not Cause

Sweden’s reading difficulties have roots that no technology decision created and no technology reversal will fix.

Radical school decentralisation. In the 1990s, Sweden transferred school governance to municipalities and opened the market to private operators (friskolor). This created substantial and growing inequality in school quality, teacher working conditions, and resource allocation. The OECD has consistently identified Sweden’s school system as among the most stratified in the OECD area. This is a structural governance problem that has nothing to do with iPads.

Teacher shortage and qualification decline. Sweden has faced a severe and worsening teacher shortage for over a decade, particularly in mathematics and Swedish. The proportion of lessons taught by unqualified teachers has risen significantly. No device policy — and no textbook purchase — addresses a structural shortage of trained educators.

Inadequate digital pedagogy — the same failure as Norway. Denmark’s former education minister Merete Riisager is quoted in the Economist admitting that “teachers were at sea” after tablet deployment: “It’s expensive to go back to books.” The Stockholm school principal Kara Barker-Astrom confirms the pattern: teacher training did not keep pace with device distribution. This is the recurring failure mode — technology deployed without the pedagogical infrastructure to use it well. The correct response is to build that infrastructure. Removing the technology is an easier political move that does not address the underlying gap.

3.3 What Edholm Gets Right

Intellectual honesty requires acknowledging the parts of Sweden’s policy that are proportionate to the evidence.

The textbook investment is defensible. The UNESCO Global Education Monitoring Report (2023) noted that structured print materials have among the highest and most consistent effect sizes of any educational intervention across contexts. The principle — that students benefit from sequential, long-form print materials with deliberate didactic design — has reasonable empirical support. SKr 2bn is an expensive way to correct for an avoidable policy mistake, but the correction itself is not wrong.

Structured dedicated reading time is unambiguously good. Cecilia Rosenbaum’s programme at Montessori Mondial Kungsholmen — 15 minutes of reading at the start of every day, books in the park, teacher read-alouds — is not anti-technology. It is evidence-based literacy pedagogy. There is no serious reading researcher who argues against dedicated daily reading time in primary school. This practice should never have been displaced by anything, digital or otherwise.

Phone bans during lessons for younger students have a growing and increasingly coherent evidence base. A Florida “bell to bell” phone ban study found significant test score improvements in year two and a substantial reduction in truancy. The mechanism is not mysterious: unstructured phone access during lessons is primarily a distraction and social anxiety amplifier, not a learning tool, for students aged 7–16. The argument for banning phones during lessons is not “screens are bad” but “this specific device, used in this specific unstructured way during class hours, creates specific and measurable harms.” That is a much more defensible claim than the broad from screen to binder framing, and it is the claim that the evidence actually supports.

The age differentiation is important and correct. Edholm’s programme applies to grundskola (ages 7–16), not upper secondary. The Stockholm school principal is right that 17-year-olds preparing for university need to learn to work with digital tools under cognitive load. Barker-Astrom’s school still switches off internet during exams — a proportionate response to a specific problem (cheating and distraction), not a wholesale rejection of digital tools.

3.4 The Pendulum Risk

The most important caveat comes from Lucia Magis-Weinberg of the University of Washington: “there is no one-size-fits-all policy.” It is also the sentence most likely to disappear in the political translation of research into policy.

A student in Swedish grundskola today will enter the workforce around 2030. The ability to work productively with digital tools, evaluate information quality in high-noise environments, and manage one’s own attention in a distraction-saturated context are not optional competencies. A schooling system that treats these as threats to be avoided rather than skills to be developed is not preparing students for the world they will inhabit.

Från skärm till pärm is an effective political slogan. It is not an educational framework. The risk is that in reacting against the overclaim of the device era (“technology transforms learning”), Swedish policy overshoots into an equal and opposite overclaim (“technology destroys learning”) — and that the pendulum, having swung past equilibrium, will require its own correction in another decade.

---

4. The Anatomy of Two Moral Panics

Looking across both cases, the pattern of error is consistent even when the substance differs.

4.1 Omitting Baseline Context

In Norway’s case: media coverage rarely mentions that Norway ranked 18th in the world, above average. In Sweden’s case: coverage rarely presents the full PISA trajectory showing that the worst decline preceded devices and the best recent performance coincided with them.

4.2 Confusing Correlation with Causation

In both countries, device deployment and PISA/PIRLS decline occurred in overlapping time windows. In both countries, other significant variables — pandemic disruption, demographic change, teacher qualification decline, systemic decentralisation — changed simultaneously. In both cases, the technology variable is selected as the explanation while others are backgrounded or ignored.

4.3 Ignoring Disconfirming Cases

The countries that outperform both Norway and Sweden in international reading assessments — Singapore, Ireland, Finland (in PIRLS), Hong Kong — all deploy digital devices extensively. This pattern is not mentioned in coverage of either country’s decline.

4.4 Mistranslating Expert Caution

Norway’s Screen Use Committee: preliminary conclusion, much to suggest, research field is falling behind. Sweden’s researchers: consistent calls for better pedagogy and teacher training alongside — not instead of — technology. The nuance of “we need more balance and better preparation” does not survive translation into either “iPads destroyed a generation” or “Sweden shows us the way back.”

4.5 The Asymmetry Between the Two Cases

The Norwegian narrative is primarily a media and political failure: a real but modest trend, catastrophised and misattributed. Norway’s own policy response has been measured — a national reading strategy focused on teacher training, library access, and reading motivation, without wholesale rejection of digital tools.

The Swedish narrative is different: it is a government-led programme built on a partially valid evidence base, but framed around a causal claim that the evidence does not fully support. This makes it more dangerous, not less — because it has institutional weight, a budget, and a political identity. When a media narrative is wrong, it can be corrected. When a policy programme is wrong at its premises, the correction is slower, more expensive, and more disruptive.

---

5. A Framework for the Countries Watching

Several European countries — Poland among them — are currently citing “the Nordic experience” as justification for restricting educational technology. Poland’s Ministry of Education’s 2026 curriculum revision has been framed in part as a response to “the Norwegian lesson.” This makes the quality of the diagnosis not merely an academic question.

The framework that the evidence actually supports is not “devices are bad” and not “devices are good.” It is more granular:

1. Teacher preparation is the critical variable. In both Norway and Sweden, the central documented failure is that teachers were not prepared to use technology pedagogically. Any technology deployment programme that does not begin with sustained, specific teacher training in digital pedagogy is deploying expensive furniture. This applies symmetrically to device rollouts and to textbook rollouts: the medium is not the message; the pedagogy is.

2. Sustained reading of long-form text requires protected time. This is a curriculum design principle, not a technology stance. Whether the threat to that time comes from fragmented screen-based reading or from poorly designed lessons is secondary. Dedicated daily reading time — structured, teacher-supported, book-based — should be a non-negotiable component of any primary curriculum.

3. Phones during lessons for younger students are a specific and separable problem. The evidence for restricting unstructured phone access during class hours for students aged 7–15 is growing and coherent. This is a specific, bounded claim about a specific harmful use pattern — not a general claim about digital technology.

4. Older students need digital competency as a curriculum subject. The ability to work productively with digital tools under cognitive load, manage attention in high-distraction environments, and evaluate information quality online are vocational necessities. They need to be taught explicitly, not assumed.

5. International comparison requires baseline context. A country ranked 18th in the world and above average has a different problem than a country ranked 40th and below average. Policy responses calibrated to one description will be miscalibrated for the other.

---

6. Conclusions

The deeper issue raised by both cases is not about Norway, Sweden, or tablets. It is about how educational communities — policymakers, journalists, school leaders, and teachers — consume, evaluate, and act on evidence in a field that claims to be evidence-based.

The PIRLS 2021 dataset is publicly available. The Ludewig et al. study is openly accessible. The Norwegian Screen Use Committee’s report exists in English translation. The PISA trend data for Sweden are on the OECD website. None of this requires specialist access or advanced statistical training to interpret at a basic level.

In a field where the stakes are the literacy of entire generations, the viral spread of demonstrably incomplete causal narratives — and the billion-kroner policy decisions that follow from them — represents a failure not of technology, but of epistemics.

The pendulum is swinging. Whether it finds a calibrated resting point, or overshoots into a new set of policy errors, will depend less on the evidence than on whether anyone in a position of educational authority is willing to read it carefully.

The most urgent digital literacy we need to cultivate is not for our students. It is for ourselves.

---

References

#	Citation
[1]	Mullis, I.V.S., Martin, M.O., Foy, P., Kelly, D.L. & Fishbein, B. (2023). PIRLS 2021 International Results in Reading. IEA/Boston College. pirls2021.org
[2]	Mullis, I.V.S., Martin, M.O., Foy, P. & Hooper, M. (2017). PIRLS 2016 International Results in Reading. IEA.
[3]	OECD (2023). PISA 2022 Results (Volume I): The State of Learning and Equity in Education. OECD Publishing. doi:10.1787/53f23881-en
[4]	OECD (2019). PISA 2018 Results (Volume I). OECD Publishing.
[5]	Ludewig, U., Strietholt, R. & McElvany, N. (2025). Reading literacy decline in Europe: Disentangling school closures and out-of-school learning conditions during the COVID-19 pandemic. Learning and Instruction. doi:10.1016/j.learninstruc.2025.102150
[6]	Skjermbrukutvalget (2023). Konsekvenser av skjerm i skolen — et kunnskapsgrunnlag. Norwegian Government.
[7]	Bjørgen, A.M. (2021). Tablets in two Norwegian primary schools. Studies in Educational Evaluation, 70, 954–965. doi:10.1080/03004279.2021.1929376
[8]	Kure, A.E., Blikstad-Balas, M. & Brevik, L.M. (2025). Digital ambitions vs. classroom reality in Norwegian lower secondary schools. Teaching and Teacher Education, 153. doi:10.1016/j.tate.2024.104843
[9]	UNESCO (2023). Global Education Monitoring Report 2023: Technology in Education — A Tool on Whose Terms? UNESCO Publishing.
[10]	Rosenkvist, M. (2023). Statement on PIRLS 2021 Norwegian results. Utdanningsdirektoratet.
[11]	World Education Blog / UNESCO GEM Report (2023). PIRLS of evidence on the impact of COVID-19 school closures on learning. world-education-blog.org
[12]	The Economist (2026, May 7). Why Swedish schools are going unplugged.

---

© 2026 Igor Wypijewski. Published at workspace.edu.pl. Open access under CC BY 4.0.