Is Jolly Phonics Effective at Improving Reading Outcomes at Scale? Finding of a Propensity Score Matching Study in Namibia

• In Event: Language, multilingualism, and literacy in global contexts

Proposal

Over the past two decades, phonics-based approaches to the teaching of reading have gained prominence across English-speaking countries (National Reading Panel, 2000; Rose, 2006; Johnston & Watson, 2005). This trend has extended to several low- and middle-income countries (LMICs) with large-scale rollouts of Jolly Phonics in Nigeria, Lesotho, Djibouti, and The Gambia. In 2022, Namibia adopted Jolly Phonics as national policy for Grade 1 instruction, marking its participation in this global movement. Despite the scale of these initiatives, robust evidence on the effectiveness of Jolly Phonics in LMICs—particularly where English is not the home language of most learners—remains limited.
This paper reports on the Namibia Jolly Phonics Study, which investigated three questions:
1. To what extent was Jolly Phonics implemented in early adoption regions?
2. How did teachers respond to the programme?
3. What impact did Jolly Phonics have on learners’ English reading precursor skills and comprehension?
Background and Literature Review:
The Namibia rollout occurred in three phases: (i) a pilot in six Oshana Region schools (2017–2019), (ii) regional implementation across Oshana (2019–2021), and (iii) nationwide adoption (2022–2025). Implementation included 2–2.5 day teacher training workshops, provision of structured learning materials, and access to the Jolly Monitor app.
International evidence suggests that structured pedagogy models—which combine professional development, learner resources, and scripted teacher guides—yield the most consistent improvements in early grade learning outcomes (Kim et al., 2018; Piper et al., 2018a; Piper & Dubeck, 2024; Cilliers et al., 2020, 2022; Fleisch et al., 2020). However, research on phonics effectiveness in second-language contexts is less conclusive. Studies from Ghana (Carter et al., 2020) and Ethiopia (Opare-Kumi, 2024) highlight advantages of mother-tongue instruction, while findings from Nigeria (Okiakor, 2024) and the Philippines (Igarashi et al., 2024) suggest benefits for early English-medium instruction. Evidence on language transfer effects is similarly mixed (Mohohlane et al., 2024; Piper et al., 2018b; Humble et al., 2024). This study contributes to this debate by providing rigorous new evidence from Namibia.
Research Design:
Taking advantage of the phased rollout, we employed Propensity Score Matching (PSM) to compare early intervention schools (treatment) with later-adopting schools (comparison). Propensity scores were estimated with a probit model using school, teacher, and learner characteristics. Four nearest neighbour matching models were tested to ensure covariate balance, with diagnostics confirming strong post-matching comparability.
The final sample included 100 schools (50 Oshana, 50 Oshikoto). Within schools, Grade 4 learners who had completed Grade 3 in 2024 were randomly selected, yielding 2,189 learners (1,078 treatment; 1,111 comparison). Data were collected in January 2025 using a Namibian-adapted EGRA alongside teacher surveys capturing demographics, qualifications, instructional practices, and Jolly Phonics exposure. Ethical protocols included informed consent from teachers and caregivers.
To evaluate the effectiveness of Propensity Score Matching, balance diagnostics were conducted. These included calculating standardized mean differences and variance ratios for covariates before and after matching. The empirical strategy combined multivariate regression analysis and Propensity Score Matching (PSM) to estimate the Average Treatment Effect on the Treated (ATT). PSM further refined the analysis by constructing a matched comparison group that closely resembled the treatment group based on pre-treatment covariates. This dual approach enhanced the credibility of the findings, approximating a quasi-experimental design.
Contribution:
This study is the first independent rigorous evaluation of Jolly Phonics at scale in the Global South. It contributes to three debates: (i) the effectiveness of structured pedagogy in LMICs, (ii) the role of phonics in second-language reading instruction, and (iii) the extent of language transfer in multilingual settings. The findings provide actionable evidence for policymakers and international partners on the viability of a popular phonics-based approach in contexts where English is not the home language.

Author

• Brahm Fleisch, University of the Witwatersrand