Is numeracy an afterthought? What Pakistan's District Data says about the Order We Teach In

In this blog post, Huma Zia Faran explores what Pakistan’s learning data reveals about the importance of early numeracy and considers how it can be better integrated into foundational learning efforts.”

“Children should be able to read before they can learn anything else.”

I have heard some version of that sentence in more meetings than I can count, usually said with confidence and almost never with evidence. It sounds obviously true: reading is how a child accesses everything else so reading must come first and numeracy can wait its turn. That assumption has quietly shaped two decades of primary grades’ learning efforts in Pakistan. However, research shows that it is not entirely true. And that what we have overlooked has important consequences for how children learn.

Policy says “both”. Practice says “reading first”.

In 2024, Pakistan notified foundational learning policies at the federal level and in Balochistan, Khyber Pakhtunkhwa, Punjab and Sindh. Each of them names literacy and numeracy as co-equal pillars. As provincial implementation plans take shape, training rosters and materials budgets (mostly project specific) taken shape, numeracy has settled into them as a parallel requirement rather than a designed instructional priority: present in the text, thin in the delivery.

The reason is inherited rather than intended. Pakistan’s biggest learning-centric primary grades intervention was the USAID-funded Pakistan Reading Project (2013-2019) that reached more than 1.7 million students and trained some 27,000 teachers. Following PRP, the Sindh Reading Program and SELECT (Sindh), PEELI and its successor EaSTE (Punjab), and a province writing an implementation plan today has twenty years of reading know-how to draw on and almost nothing equivalent for numeracy.

The result is an implicit hierarchy. Literacy arrives through structured interventions, purpose-built materials and teacher support. Numeracy is left to the textbook and to whatever a teacher can improvise.

The problem, to be clear, is not that literacy receives too much attention. It is that numeracy receives too little and the assessment data shows what that costs. The National Achievement Test 2023, covering a representative sample of over 23,000 students, found Grade 4 students answering only 49 per cent of maths items correctly, against 56 per cent in English and 68 per cent in Urdu. By Grade 8, maths falls to 42 per cent: these gaps do not resolve with time, they harden. ASER 2023 adds household-level confirmation: 46 per cent of Grade 5 children could complete a two-digit division task expected of a Grade 3 student, against 50 per cent reading an Urdu story and 54 per cent reading English sentences. In TIMSS 2019, Pakistan’s first and only participation, 73 per cent of students failed to reach even the low international benchmark in maths.

So we asked a different question

Rather than asking again how badly children perform in maths, our new policy brief, Is numeracy an afterthought? (July 2026), asks whether early maths matters for everything else. Using district-level ASER data, Grade 1 scores from 2019 were matched to Grade 3 and Grade 5 outcomes from the same districts in 2023 using a series of ordinary least squares regressions.

Grade 1 maths turns out to be a statistically significant predictor of every later outcome tested. It predicts Grade 3 maths (β = 0.266) and Grade 5 Urdu/Sindhi (β = 0.266). But the striking result is what it does for English. Grade 1 maths predicts Grade 3 English at β = 0.323, explaining 10.4 per cent of the variance; by Grade 5 that rises to β = 0.380 and 14.5 per cent.

Two things here deserve a pause.

First, at Grade 3, early maths out-predicts early English for English itself. Grade 1 English explains 8.6 per cent of the variance in Grade 3 English (β = 0.294); Grade 1 maths explains 10.4 per cent. A child’s early number sense tells you more about their Grade 3 English than their early English does.

Second, and I want to be straight about this because the honest version is the more useful one. By Grade 5, early English reclaims the strongest single association with Grade 5 English (β = 0.420, R² = 17.8 per cent). Language-specific skills reassert themselves as literacy demands grow, which is exactly what one would expect. What does not fade is maths. Its explanatory power on English accelerates, growing by roughly half between Grade 3 and Grade 5. Early numeracy behaves less like a competing subject and more like a compounding asset and the full head-to-head estimates are set out in the accompanying technical note.

These are district-level associations, not individual-level causal estimates, and the technical note says so plainly, alongside the multicollinearity among Grade 1 subject scores (r = 0.85–0.88) that pushed the analysis from a multivariate specification to independent OLS models. But the direction is consistent with a substantial international literature: Duncan and colleagues (2007) across more than 35,000 children, Ritchie and Bates (2013) on links running all the way to adult socio-economic status, and a recent meta-analysis of 3,308 participants (Ünal et al., 2025) showing that reading and maths draw on a shared cognitive-control network that children engage more extensively than adults do.

An implementation problem, not a policy one

Pakistan does not need another foundational learning policy. It has five. What it lacks is an operational answer to four questions: how numeracy skills should be taught, how much protected instructional time they require and how teachers will be supported to deliver them.

Without those answers, numeracy will be delivered the way it has always been delivered: procedurally, inside combined learning hours, judged by syllabus completion. That is precisely how a co-equal pillar quietly becomes the unimplemented one.

The brief proposes four corrections that federal and provincial departments could make immediately, using existing structures and existing money:

1. Protect the time. Issue an implementation directive requiring every school to earmark dedicated daily time for “morning maths” block for number sense, estimation, pattern recognition and reasoning, when children’s attention is highest, and verify it through routine school monitoring.
2. Make it visible. Declare 2026-27 the Year of Foundational Numeracy, with weekly early-grade activities, district-level challenges, parent-facing practice materials and monthly leadership messaging. Pakistan has run reading campaigns for a decade. It has never run one for maths.
3. Ring-fence what already exists. Require a stated share of every foundational learning investment i.e. teacher training hours, instructional material packages, assessments, technical assistance to be explicitly allocated to and reported for numeracy. This is a reporting change, not a financing ask.
4. Monitor teaching, not testing. Observe whether numeracy routines, planned activities and basic manipulatives are actually in use. Additional student testing distorts classroom practice; observing instruction does not.

The longer-term work is harder: curricula that value depth over coverage, reasoning-rich pedagogy, formative assessment designed to inform instruction rather than to certify or select (as set out in our Foundational Learning Assessment Framework), and teacher development built on sustained coaching rather than one-off workshops (Reimagining CPD). Sustained political attention matters too, such as an annual “State of Numeracy” report to standing committees of the National and provincial assemblies would outlast any single government or donor cycle.

Why this travels beyond Pakistan

None of this is a Pakistani peculiarity. Across much of the global South, the foundational learning agenda has been operationalised through starting with reading. Numeracy was written into projects and/or policies, and then handed to business as usual. Yet Hanushek and Woessmann have long shown that a country’s stock of quantitative cognitive skills predicts long-run growth better than years of schooling, most strongly in developing economies, and Pakistan’s own employers have been saying the same about school leavers for years.

Singapore’s concrete-pictorial-abstract sequence, Finland’s play-based number sense and the UK’s teaching-for-mastery approach differ enormously in method. They agree on one thing: numeracy is treated as a core foundational competency from the earliest years, not as a subject that arrives once reading is settled.

So the question I would put to this network is not whether literacy matters. It plainly does, and nothing here argues for less of it. The question is narrower and more uncomfortable: in your context, is numeracy actually resourced as a co-equal pillar of foundational learning or is it merely named as one? And if it is only named, what would it take, in the next twelve months and without new money, to change that?

I would genuinely like to hear where the answer differs from ours.

About the author

Huma Zia Faran leads policy research at the Pak Alliance for Maths and Science in Islamabad, where she works on foundational learning, assessment and teacher development, and is the author of the July 2026 policy brief Is numeracy an afterthought? A necessary course-correction in Pakistan’s push for foundational learning, on which this post is based. She can be reached at huma@pams.pk

References

ASER Pakistan. (2023). Annual status of education report – National. https://aserpakistan.org/document/2024/aser_national_2023.pdf

Duncan, G. J., Dowsett, C. J., Claessens, A., Magnuson, K., Huston, A. C., Klebanov, P., Pagani, L. S., Feinstein, L., Engel, M., Brooks-Gunn, J., Sexton, H., Duckworth, K., & Japel, C. (2007). School readiness and later achievement. Developmental Psychology, 43(6), 1428–1446. https://doi.org/10.1037/0012-1649.43.6.1428

Hanushek, E. A., & Woessmann, L. (2007). Education quality and economic growth. World Bank. https://hanushek.stanford.edu/sites/default/files/publications/Hanushek%2BWoessmann%202007%20Education%20Quality%20and%20Economic%20Growth.pdf

Pak Alliance for Maths and Science. (2025a). Foundational learning assessment framework. https://mathsandscience.pk/publications/fl-assessment-framework/

Pak Alliance for Maths and Science. (2025b). Reimagining CPD. https://mathsandscience.pk/publications/reimagining-cpd/

Pak Alliance for Maths and Science. (2026a). Is numeracy an afterthought? A necessary course-correction in Pakistan’s push for foundational learning. https://mathsandscience.pk/wp-content/uploads/2026/07/Is-numeracy-an-afterthought-Policy-Brief.pdf

Pak Alliance for Maths and Science. (2026b). Technical note: Is numeracy an afterthought? https://mathsandscience.pk/wp-content/uploads/2026/07/Technical-note-Policy-Brief.pdf

Pakistan Institute of Education. (2023). National achievement test 2023 findings report. https://pie.gov.pk/SiteImage/Publication/NAT%202023.pdf

Ramírez, M. J., Mahmood, S., Khan Hamid, F., & Shakeel, M. (2019). TIMSS 2019 national report Pakistan: How much are our fourth-graders learning at school? IEA. https://www.iea.nl/publications/study-reports/national-reports-iea-studies/timss-2019-national-report-pakistan

Ritchie, S. J., & Bates, T. C. (2013). Enduring links from childhood mathematics and reading achievement to adult socioeconomic status. Psychological Science, 24(7), 1301–1308. https://doi.org/10.1177/0956797612466268

Ünal, Z. E., Park, Y., Simsek, E., Menon, V., & Geary, D. C. (2025). Neurodevelopmental commonalities in the cognitive control networks for mathematics and reading. Nature Communications, 16, 8398. https://doi.org/10.1038/s41467-025-63259-8