(Revised version 10/09/2025)
The Wall Street Journal recently published a letter by Tim Shanahan in response to my earlier letter on phonics instruction in California. Several of Shanahan’s points are wrong or misleading.
1. Shanahan describes Jeffrey Bowers’s 2020 study—which concluded that systematic phonics is no more effective than other forms of reading instruction—as “wanting” and “seriously flawed.” He also states that Bowers’s “main point” is that we should implement a different type of phonics instruction.
These two statements are difficult to reconcile. If the study were methodologically unsound, its conclusions—whatever they may be—would not be meaningful.
In fact, Bowers’s paper, “Reconsidering the Evidence That Systematic Phonics Is
More Effective Than Alternative Methods of Reading,” has as its main point precisely what is specified in its title. It’s an survey of various meta-analyses and other data that have been put forth to claim the superiority of explicit, systematic phonics instruction over whole language and other methods.
Bowers’s suggestion that the field should explore an alternative instructional approach he favors appears only once (although the approach is briefly described earlier in the paper). That suggestion is a single 48-word sentence in the final paragraph of a 24-page paper exceeding 14,000 words:
“For example, one possibility is that reading instruction in English should focus more on the role that meaning plays in organizing spellings (via morphology) and that English spelling system makes sense once the interrelation between phonology, morphology, and etymology are considered (Bowers and Bowers 2017, 2018c). Of course, other possibilities need to be considered as well, but the first step in motivating more research into alternative forms of instruction is to realize that there is a problem with the current approach.” (Bowers, 2020, p. 703)
That statement does not alter the empirical findings reported in the paper, which were summarized accurately in my letter.
I’m baffled as to how Shanahan, an experienced researcher, could consider Bowers’s singe sentence as the “main point” of the article unless he read only the final paragraph of the piece. (Bowers’s alternative is actually not a “phonics” approach, so even this statement is inaccurate.)
2. Regarding the quality of the study, three formal critiques of Bowers (2020) have appeared. In only one case was Bowers given the opportunity to respond in the same journal, a breach of normal academic protocol. His published reply and additional responses on his website address the specific criticisms in detail. Readers may review those exchanges directly. In my view, Bowers completely eviscerates the critiques of his study (here, here, and here).
3. Shanahan agrees (contra the WSJ’s original editorial) that California adopted phonics-based instruction in the 1990s, as I stated in my letter. But he also states that 4th-grade reading scores increased substantially afterward. This claim is only true if we stretch the evidence beyond the breaking point and include the 25-year period after the first mid-1990s reforms were enacted.
California’s 4th-grade NAEP reading score was 202 in 1998 (prior to full implementation of the phonics reforms). By 2002 it had risen to 206, a difference that was not statistically significant. In 2009, approximately 15 years after the reforms began, the score was 210—an improvement relative to 1998 but not meaningfully different from 2002.
Shanahan notes that scores rose 14 points “pre-pandemic” and that California eventually “tied” the national average. This is false. The 14-point gain by 2019 is correct; however, California remained below the national mean at that time. The statistical “tie” occurred only in 2022, after the national scores plummeted post-pandemic.
Even if we were to grant Shanahan’s generous time frame for the effects of the 1990s phonics reforms, doing so creates another problem for his argument. If systematic phonics instruction has been in place in California for nearly three decades, why do we need additional measures of the same type now?
Advocates of the “science of reading” have argued that phonics was later replaced by “balanced literacy” sometime in the early 2000s. But if that is true, and if the state’s reading scores improved during the subsequent period in the 2010s, then all of those gains would have been due to the much-reviled “balanced literacy” approach rather than through phonics alone.
4. Shanahan attributes improvements in early reading achievement to phonics but cites no direct evidence supporting that claim. As noted, California’s scores did not rise for 15 years after adopting systematic phonics. Mississippi’s 4th-grade scores did increase, but the state simultaneously implemented several other policies, including retention of 3rd-grade students, making it difficult to claim phonics as the main reason for the improvement in scores.
5. For the record, I don’t think we should be talking about any NAEP state scores at all in this debate. My letter used California and Mississippi data because that’s what the Journal editorial relied on. But state reading programs aren’t experiments and provide the lowest-quality evidence we have. We should instead rely on carefully controlled, rigorously analyzed experiments of the sort that Bowers reviewed. And that evidence, even in its most optimistic light, shows very weak support for an explicit systematic phonics approach.
6. The remainder of Shanahan’s letter concerns reading instruction for older students. He recommends in a recent blog post interventions focusing on vocabulary, comprehension strategies, and fluency. However, recent meta-analyses have found that these approaches do not reliably improve general reading comprehension (see here, here, and here).
Follow Us!