In the past two decades, something extraordinary has happened with very little fanfare: The reading ability of our lowest-performing children has increased significantly. According to the National Assessment of Educational Progress (NAEP), between 1990 and 2012, the scores of nine-year-olds at the tenth and twenty-fifth percentiles increased by roughly two grade levels (about twenty points). For those children, those gains aren’t just impressive—they’re potentially life-changing.
At the same time, there has been a fourteen-point gain (a little more than a grade level) among fourth graders at the fiftieth percentile and a mere six-point gain among those at the seventy-fifth and ninetieth percentiles.
What’s causing this long-term trend of much greater gains among lower-performing students than higher-performing ones? That’s hard to say. There are many plausible explanations, but one that seems likely is that K–2 teachers have simply gotten better at teaching “decoding” (learning to sound out words). Nationwide, there’s been an increased focus on evidence-based practices, including high-profile initiatives like the National Reading Panel report and Reading First. Both stressed that children must be explicitly taught how to decode, and most early reading programs—and, more significantly, teachers—seem to have gotten the message.
But decoding is only the first step in developing strong readers. And there’s reason to believe that in striving for proficiency, many other steps have been neglected. As Nell K. Duke, one of the nation’s reading researchers, wrote:
When the aim is to show reading improvements in a short period of time, spending large amounts of time on word-reading skill and its foundations, and relatively little on comprehension, vocabulary, and conceptual and content knowledge, makes sense. Measurable gains in phonological awareness, alphabet knowledge, and word reading can be achieved quickly and, for most students, relatively easily. In contrast, gains in comprehension, vocabulary, and conceptual knowledge are harder to measure, at least in young children, and harder to achieve. Yet the long-term consequences of failing to attend to these areas cannot be overstated [emphasis added].
Unlike decoding, reading comprehension is not a skill that can be directly taught, practiced, and mastered. Expansive knowledge and vocabulary, accumulated over time, are the keys to comprehension. If you know a lot about dinosaurs, for example, and have learned to decode, you can read with understanding about dinosaurs with little difficulty. But if you’ve never heard of dinosaurs, your ability to read complex texts about them is compromised.
As your broad, general knowledge grows, so will your broad, general reading ability. The sheer volume of subjects a literate individual has some knowledge of, and the size of a proficient reader’s vocabulary, mean that the race to improve reading comprehension is by definition a marathon. But our testing and accountability policies have tended to demand that teachers treat it as a sprint. That might be why NAEP reading scores among thirteen-year-olds have only increased by 3–8 points (looking across the spectrum from the tenth to the ninetieth percentiles) between 1990 and 2012; among seventeen year olds, they have decreased by 3–5 points. These are the long-term consequences to which Duke alludes.
By returning control to the states, the Every Student Succeeds Act (ESSA) provides an opportunity for states to rethink the enticements baked into accountability policies. They can now incentivize long-term investments in building knowledge and vocabulary over short-term investments in boosting scores, which too often spike quickly before plateauing or even fading.
It is not a mystery why reading comprehension scores are so stubbornly tied to socioeconomic status. Knowledge and vocabulary grow exponentially, beginning at birth. Children with well-educated parents come to school with larger vocabularies and more knowledge. Their verbal advantage grows each day during dinnertime conversation, bedtime read-alouds, weekend museum visits, sports and music lessons, and other forms of “concerted cultivation.” Reading researcher Keith Stanovich dubbed this structural advantage in language and knowledge the “Matthew Effect,” from the biblical passage in the Gospel of Matthew about the rich getting richer and the poor getting poorer. It neatly captures the heart of the issue: Privileged children come to school on day one with larger stores of knowledge and vocabulary than their disadvantage peers. That gives new knowledge and vocabulary fertile soil in which to root. The gaps don’t merely persist, they widen. The longer we wait, the wider the gap grows.
Valorizing knowledge acquisition is the secret sauce that’s missing from education policy, testing, and accountability. Preferred policy areas—like teacher quality, choice, chartering and merit pay—are agnostic to curricular content. This is a hiding-in-plain-sight lever that policy makers have seldom thought to pull. ESSA could change that.
The principal challenge for state policy makers is to embrace a clearer, more accurate view of what reading comprehension actually is: a reflection of the sum of a child’s education across the curriculum, not a “skill” to be taught. Once you see reading through this lens, a different policy picture starts to emerge almost unbidden. What kids do in school all day—not just in reading instruction—starts to matter a lot. If this insight were reflected in the way we structure and incentivize schools, elementary school would look very different than it does today. To pick one obvious example: A nationally representative survey published in 2012 found that K–3 teachers spent just 16–19 minutes per day on social studies and science; grades 4–6 teachers spent just 21–24 minutes a day. This is exactly the wrong approach once you see building knowledge as integral to reading achievement.
“The mistaken idea that reading is a skill,” notes cognitive scientist Daniel Willingham, “may be the single biggest factor holding back reading achievement in the country. The knowledge base problem must be solved.”
If we want all children to be great readers, education policies must actively incentivize teaching reading skills plus science, history, geography, and the arts from the first days of school. Substantial gains in decoding have shown we can get kids to the starting line. But we’re leaving them stuck there.
Next week, we’ll talk about the implications for state policy makers of this clear view of what reading comprehension actually is.