Representation and Absence: Race and Ethnicity in the World’s Best-Selling ELT Coursebook

Abstract

Claims about racial and ethnic representation in ELT materials are frequently made without specifying the demographic standard against which representation is being judged. As a result, the same data may support different conclusions. This study applies three explicit benchmarks — United Kingdom publisher-country demographics, global population demographics, and global English language learner demographics — to the six Student Books of the Headway 5th edition, a series marketed as the world’s best-selling adult English course. Every photographed human figure whose racial or ethnic identity was visually legible was counted and assigned to one of eight categories. Of 1,877 photographed figures, 1,823 were codeable and 54 were recorded as ambiguous and excluded from benchmark comparisons. A second coder independently recoded a systematic sample of 197 pages (20.2% of the corpus), identifying 414 figures against the primary coder’s 415, with 98.3% category-overlap agreement. White figures account for 78.2% of the codeable corpus and Black figures for 9.0%. The three benchmarks yield substantially different readings of these results: against United Kingdom census data the White share sits marginally below the national proportion and the Black share at more than double it, while against global population and global learner benchmarks the White share is several times larger than any corresponding demographic proportion. Southeast Asian figures account for 1.2% of the codeable corpus, and no figure in the series was coded as Central Asian, despite Southeast and Central Asian populations forming a substantial and growing share of English language learners worldwide. The study argues that specifying demographic benchmarks is a precondition for interpretable claims about representation, and documents a pattern of visual absence that a Black/White framing of diversity does not capture.