Get stories like this delivered straight to your inbox. Sign up for The 74 Newsletter

A version of this essay originally appeared on Robert Pondiscio’s Substack.

The most recent National Assessment of Educational Progress (NAEP) results delivered a familiar gut punch: Just 30% of eighth graders read at or above the proficient level, a number that’s barely budged in decades. Even in states like Mississippi and Louisiana, which have earned national attention thanks to literacy reforms that have smartly lifted fourth-grade scores in recent NAEP cycles, early gains tend to plateau or evaporate by eighth grade. A substantial number of U.S. students simply seem to run out of gas as readers as they move from upper elementary to middle school and beyond. 

A compelling explanation may lie in something called the decoding threshold. Teachers often assume that once students master decoding in early elementary school, they’re set to shift from learning to read to reading to learn. However, in 2019, researchers at the Educational Testing Service published a noteworthy study that measured foundational literacy skills — like decoding — in students from upper elementary through high school. Most reading tests in the older grades focus solely on comprehension; they don’t offer much insight into whether students have mastered the basic skills necessary to read fluently. The findings showed evidence of a troubling phenomenon: Students with weak decoding skills consistently performed poorly on comprehension tasks, while those who surpassed a certain level of decoding ability tended to understand texts much more effectively. In other words, although decoding isn’t the only skill older students need to succeed in reading, those who haven’t yet mastered it are likely to struggle with understanding complex material.

A follow-up study three years later confirmed it: Those below the decoding threshold stagnated, while those above the line advanced — offering tantalizing evidence to explain why eighth-grade NAEP scores plateau even as fourth-grade numbers rise. A recent research brief put the matter succinctly and starkly: “If children do not have adequate word-recognition skills, their reading comprehension often won’t get better no matter how much direct support for comprehension they receive.” 

The tripwire that appears to be holding kids back is multisyllabic decoding. Students who can decode simple words like “cat” and “bed” with relative ease may still struggle to break down longer, more complex words into smaller, manageable parts to read them correctly. Imagine two eighth graders reading a science passage that includes the word “photosynthesis.” The student above the decoding threshold effortlessly breaks it into “photo” and “synthesis,” adjusts the sounds in her head — like “syn” to “sin” — and reads it smoothly, quickly grasping it as a plant process she’s studying. Meanwhile, the student below the threshold freezes at the unfamiliar term and mangles it as “photo-sith-esis” or “photo-sy-thee-sis.” Struggling to decode the big word, he loses the thread of the sentence, missing the whole idea of plants making energy. 

It’s another manifestation of cognitive load theory: Brainpower spent decoding multisyllabic words is not available to attend to the meaning of the text. Worse, the decoding threshold fuels a rich-get-richer, poor-get-poorer phenomenon often referred to as the Matthew Effect: Students who are below the decoding threshold stop growing in vocabulary, reading comprehension and knowledge acquisition, while those who are above have what it takes to keep learning and growing, leaving the struggling readers in their wake. 

Worse still, evidence of the decoding threshold reveals a blind spot in common approaches to teaching reading. “We basically don’t teach [multisyllabic decoding] anywhere in the system because it’s too advanced for second graders. And after second grade, we stop decoding instruction and flip into comprehension and fluency,” observes Rebecca Kockler, a former Louisiana state education leader who now heads Reading Reimagined, a $40 million initiative of the Advanced Education Research and Development Fund. “If I had a magic wand, I would pull decoding fluency work up almost into seventh or eighth grade,” she says, while pushing down to early elementary grades the building blocks of multisyllabic decoding, such as morphology and etymology. If you teach kids to break words into their smallest meaningful pieces, like “un-” for “not” or “-ness” for a state of being, they’re more likely to be able to handle “unhappiness” by spotting its parts, for example. And by showing them where words come from — like how “photo” in “photosynthesis” means “light” from Greek — they will be better able to infer what words mean.

As persuasive as the decoding threshold thesis might be, the wish for a magic wand to wave at curriculum and standards hints at a serious problem: There is no immediate or obvious solution at hand to address the issue. Nor is there simply a lack of appropriate curriculum or materials. A recent RAND survey of teachers in grades 3 to 8 found that 44% of their students “always or nearly always experience difficulty” reading the content of their instructional materials. The report also found many of those same teachers hold misconceptions about how students develop word recognition skills.

A new nonprofit venture called Magpie Literacy, a collaborative effort with the fund led by Kockler, has been piloting a set of tech-enabled instructional tools aimed at addressing these issues directly. In a 12-week pilot in grades K-2 across 11 schools in Little Rock, Arkansas, early results were promising, with evidence of impact in only 8 to 12 weeks of use. Student growth was most pronounced, according to Kockler’s colleagues, among students starting at the lowest levels of proficiency. K-2 may sound early to address a problem that shows up starkly in eighth grade, but it reflects a growing conviction: unless students start building sophisticated decoding skills young, and those skills are reinforced often, too many will continue to hit the wall in middle school and never get back up to speed. “We’ve had this belief that we teach kids to read and then they read to learn,” Kockler explains, “and we just fundamentally do not believe that’s true anymore.” 

If you had asked her years ago, when she was assistant superintendent of academics with the Louisiana Department of Education, to estimate the percentage of middle schoolers who struggled with decoding to the point that it interfered with their reading comprehension, Kockler would have guessed 7% to 10%. “We think that number is more like 30% to 40%,” she now says, “which really mirrors this group of middle schoolers who never ever show growth on state tests or NAEP.” 

Get stories like these delivered straight to your inbox. Sign up for The 74 Newsletter