Why Aren't American Reading Scores Higher?

  • 19 January, 2019

Blast from the Past: Four years ago, I posted this blog. These days -- post COVID -- this question is coming up even more often, and more insistently than then. A lot of instruction was lost by many children and schools are scrambling to do better. Unfortunately, too many are seizing on simplistic solutions, that might be helpful for some kids, but which will not be likely to improve achievement very much. The solutions provided in this blog are still the best way to go for those who are serious about raising reading achievement -- in their classroom, in their school, or in their state or school district.

Teacher question:

The National Assessment of Educational Progress says that only 37% of 4th graders are reaching reading proficiency. Why is it so low?

Shanahan response:

Why do so few American kids read well?

There seems to be plenty of blame to go around. Parents? Society? Too much screen time? Poverty? Immigration? You and me?

Lots of possibilities; some truth to each of these explanations. We’re all responsible, none of us are responsible. Yada, yada.

Poverty is a big problem, of course. The correlations between test performance and family income (or family security) are high. Childhood poverty can affect the brain—making it less able to learn (Noble, 2017)—and, SES includes not just income but also parents’ education. Parents who are less likely to read and to read well are less likely to raise children who thrive in school. Poverty also contributes to poor nutrition, school-interrupting illnesses, and stress. There are exceptions, of course, but the patterns are strong. American schools serve a lot of children who live in poverty and that contributes big time to low achievement (high income schools score well, high poverty schools do not).

It is true that children from low income schools have a higher likelihood of struggling in school, but it is also true that some schools manage to address these children’s educational needs better than others. At least some other countries manage to deliver more effective education to their poor kids (Darling-Hammond, 2014).

I suspect that one of the things that makes explanations like poverty or uncaring parents or high screen time so attractive to so many educators is that none of those are problem that we can address.

It’s the equivalent of the waiter who tells you, “Not my table,” when you ask for help and your own waiter has gone missing.

“I could teach these children successfully if they weren’t poor or if dad hadn’t walked out or mom spoke English or if the kids weren’t already so far behind by the time that they reach my class…” Sounds committed and hopeful and positive—if someone would just fix these kids, I’d be able to teach them…. But, obviously, schools can’t “fix” the kids, so the hopefulness deteriorates into so few children reading proficiently.

I have no doubt that those with different expertise than mine should be trying to address issues of poverty and racism and child health and anything else that makes learning harder. But these are the children we have, and they are at our table. Our job is to serve their needs as well as possible. 

What can educators do to raise achievement?

It’s complicated but I think that it has to do with opportunity to learn. We simply don’t spend enough time on those things that make a difference in making kids proficient. Most American elementary schools these days pride themselves on their 90-minute reading blocks… but much of that time gets devoted to things that do little to promote children’s reading ability:  the kids are supposedly reading on their own or doing keep-busy-but-keep-quiet sheets while the teachers are working with other kids.

I’d love it if instead of a 90-minute block, we’d commit to providing 90 minutes of teaching and guided practice to each child each day. That might take more than 90 minutes to accomplish, but it would sure give kids a better chance to become proficient. (The 90-minute block is often a myth anyway; watch closely and you’ll see that reading lessons don’t actually begin at the beginning of the school day and yet those school opening minutes are counted as literacy block time.)

In my schools, I required 120-180 minutes per day of reading and writing instruction. I know that’s a lot, but it is accomplishable in most schools (and if you get rid of test prep and extended specific skills teaching in reading comprehension and other things that don’t enable students to read better, meeting those times can even be easier).

This instructional time should be devoted to explicit teaching and guided practice aimed at developing knowledge of words (including phonemic awareness, phonics, letter names, spelling, morphology, vocabulary); oral reading fluency; reading comprehension; and writing. And, for English learners (and perhaps poverty kids too)—explicit oral language teaching.

Too many American teachers have bought into the idea that kids would be better off reading on their own than working with the teacher because reading is learned by reading. I agree with the idea that reading matters in learning to read, but such reading is better included in the reading lessons than pushed away from the lessons. I’ve argued that at least half of the instructional time (perhaps more) should be spent reading and writing.

That means in a reading comprehension lesson, there will be teacher-led demonstrations and explanations, and guided discussions, and so on—but the kids would be reading throughout these activities. The same is true for decoding instruction; a big chunk of that time should involve the kids in decoding and encoding words.

Second- and third-graders spend too much time working only with books they can already read reasonably well—and that idea has been spreading up the grades. Despite the claims of some educators and marketers, there is no such thing as an instructional level in reading (at least beyond the very beginnings of reading). Teaching kids at their supposed “reading levels” hasn’t been found to facilitate learning, but it does lower the sophistication and complexity of the content and language kids are working with.

I suspect that most teachers do little to support or extend students’ reading stamina. Oh, I know that some are proud that they use books instead of short stories to teach reading, or that many assign extended silent reading. But those tend to be sink-or-swim propositions. To make extended reads successful many teachers walk the kids through the texts round robin-style or have the kids read short sections of text interspersed with discussion or teacher explanation. Kids would be better prepared for tests (and many real reading situations) if there was an intentional regimen of stretching how long they can persist in making sense of texts. For a lot of kids, when they have to read an entire fourth-grade selection silently to answer questions about it as on the NAEP assessment, it doesn’t go so well since they’ve never done anything that demanding before.

Lack of a knowledge-focused curriculum is an important culprit, too. Science and social studies aren’t given enough time in elementary school (and the value of the literature may be suspect, too). We need to provide daily teaching in these other subject areas, and those lessons should include the reading of rich content text. Such texts also should be able to find a place within reading instruction, too. I’m a big fan of including content learning objectives in reading programs.

(Note to Educational Policymakers: This statement is aimed at teachers. But that does not get you off the hook. Providing policies and resources that ensure that teachers have sufficient professional development and support and sufficient amounts of teaching time are on you.)

Nothing very exciting here, right?

If we want to more of our kids to be reading proficient at the levels needed in the 21st century, it will take a lot of dedicated teaching of the key things that matter in learning. Nothing sexy about it, and yet too few kids get those things and placing blame doesn’t help. Focus like a laser on what works rather than on what you like to do and these kids are likely to do better. 


See what others have to say about this topic.

Faith Gerber May 02, 2019 07:18 PM

As always, straight to the heart of the matter. I just finished writing my master's thesis on disciplinary literacy in middle school, and research confirms the need for more elementary social studies and science instruction that includes lots of literacy.

Colin Smith Dec 14, 2020 12:26 AM

Hello Faith,

It does little good to write an article while completely missing the bigger picture. For one, it’s unclear if poverty is truly affecting test scores. As Dr. Sandra Scarr writes, “what the public deems impoverished differs from experts.” Meaning, that rampant child abuse, malnutrition, and emotional disinvestment can stunt neurological functioning. However, that is less than 2% of all school-aged children. More importantly, test scores have stagnated amongst all groups and socioeconomic classes — except the highest scorers. Your analysis was not thorough enough and failed to answer why test scores are falling throughout each ethnic group.

What is occurring correlates to the Bell Curve and a reversing Flynn Effect. We have peaked regarding the ability of students to meet standardized testing goals. Therefore, education is showing diminished returns insofar as more funding and better instruction cannot raise test scores. What is necessary to change the trend is recalibrating the goals for each child’s unique ability. For instance, a child with an IQ of 85 will likely never score on grade-level for reading. Is the issue the school, teacher, or statewide tests? None of the above; one could argue that the child is scoring adequately for their individual competency. Also, do not discount cognitive stratification and a dysgenic trend downward. The children today within impoverished schools are less genetically intelligent as natural selection reverses cognition over time.

Kathleen Leos Apr 22, 2023 02:28 PM

It’s all about the research and pd and instruction or input based on the esprit reading research.

Jeanne Apr 22, 2023 02:54 PM

Kathleen Leos
I'm not clear on your meaning. Can you be more specific and elaborate as well?

Tom Gunning Apr 22, 2023 03:17 PM

We can double the number of children who read well by using the word "proficiency" correctly. After studying NAEP levels, a distinguished group of assessment scholars concluded, “ . . . there is no mention of ‘at grade level’ performance in the achievement levels and, hence, the Proficient level does not reflect ‘at grade’ performance nor is it synonymous with ‘proficiency’ in the subject; the Basic level is less than full mastery but more than minimal competency . . . .” (National Academies of Sciences, Engineering, and Medicine, 2017, p. 246). The proficient and advanced levels are above grade level (Loveless, 2016; Pellegrino, Jones, & Mitchell, 1999). In a 2018 article Peggy Carr, now commissioner of the National Center for Education Statistics (NCES), stated:
We have gotten an earful from our stakeholders about the confusion of the word “proficient” as we use it. . . . NAEP sorts students’ scores into three achievement levels: basic, proficient, and advanced. And to understand how many US students are performing on grade level, it is far more accurate to look at “basic” scores, rather than “proficient.” (Hinckley, 2018, April 12).

Tom Gunning Apr 22, 2023 03:18 PM

We can double the number of children who read well by using the word "proficiency" correctly. After studying NAEP levels, a distinguished group of assessment scholars concluded, “ . . . there is no mention of ‘at grade level’ performance in the achievement levels and, hence, the Proficient level does not reflect ‘at grade’ performance nor is it synonymous with ‘proficiency’ in the subject; the Basic level is less than full mastery but more than minimal competency . . . .” (National Academies of Sciences, Engineering, and Medicine, 2017, p. 246). The proficient and advanced levels are above grade level (Loveless, 2016; Pellegrino, Jones, & Mitchell, 1999). In a 2018 article Peggy Carr, now commissioner of the National Center for Education Statistics (NCES), stated:
We have gotten an earful from our stakeholders about the confusion of the word “proficient” as we use it. . . . NAEP sorts students’ scores into three achievement levels: basic, proficient, and advanced. And to understand how many US students are performing on grade level, it is far more accurate to look at “basic” scores, rather than “proficient.” (Hinckley, 2018, April 12).

Lauren Apr 22, 2023 03:24 PM

Are there any studies out there about the tests themselves over the past decades? How have the state assessments changed, and is this part of the equation as well? I work with students who can read a grade level book and discuss the meaning demonstrating comprehension, but when it comes to state tests, they don't do well. Does the online aspect confuse them? Does the format of the test confuse them? What are the tests measuring? Is it strictly comprehension? Are there other specific concepts and vocabulary that students need to know to do well on the tests such as: Author's purpose, personal connection, point of view, simile, analogy... When we compare test results from the past with current test results, are we comparing apples to oranges?

Andrew Biemiller Apr 22, 2023 03:38 PM

Hello Tim,
The continuing losses in reading comprehension reflect two problems--lack of adequate skill to decode printed language (including sufficient practice), and inadequate vocabulary (and associated knowledge). The strongest correlate of later (post grade 4 reading comprehension) reading is vocabulary.
Michal Shany and I demonstrated that reading achievement could be improved with half an hour's "assisted reading" for grade 3 and 4 "poor readers'" reading practice. The UPPER LIMIT for children's gains was vocabulary (PPVT).
We could have a lot of progress to be made by building stronger vocabularies. (As you observe, his would come from richer reading as well as vocabulary acquisition skills and a small addition of direct vocabulary instruction.)

Shany, M., & Biemiller, A. (1995). Assisted reading practice: Effects on performance for poor readers in grades 3 and 4. Reading Research Quarterly, 30, 382-395.

Shany, M., & Biemiller, A. (2009). Individual differences in reading comprehension gains from assisted reading practice: Pre-existing conditions, vocabulary acquisition, and amounts of practice. Reading and Writing, 23, 1071-1083.

Cheers, Andy Biemiller

Susan Van Zant Apr 22, 2023 03:48 PM

I am so glad that you mentioned the need for professional development. Many teacher are well meaning; however, they lack knowledge related to systematic and explicit teaching of the basic foundational skills. When these teachers participate in professional development related the science of reading they better understand the reasons for the format of their adopted reading programs. The teachers are thus better able to meet the needs of all students, especially those who struggle with reading, spelling and writing. They are willing to devote 180 minutes to reading. Throughout the rest of the school day these teachers often point out reading patterns in words found in math, social studies and science. As a result their students become more capable readers and have higher achievement scores on state tests.

As an aside: I was in the audience a while back. You mentioned that you had been a machinest and then became a tool and die maker. Previous to becoming a educator I worked for the Department of Employment. I knew that a machinist was a good job...but a tool and die maker was a step up. You would have to know how to read blue pirnts, know/apply higher math skills and use many different machines, etc. When you mentioned this I looked around the room to give a "thumbs-up" to others. Evreyone in the room just kept listening or taking notes. It was at time I realized the importance of prior knowledge. I was impressed because I understood the difference in these two jobs, and they did not.

Harriett Janetos Apr 22, 2023 03:55 PM

Here's what baffles me: We have known about the best way to teach vocabulary since Ehri and Rosenthal's 2007 research, Spellings of Words: A Neglected Facilitator of Vocabulary Learning, and yet I still see recommended activities that don't unite phonology, orthography, and semantics. This is part of the problem. From the abstract:

"Teachers need to show the spellings of new vocabulary words when they discuss their meanings. Students need to stop and pronounce unfamiliar words rather than skip them during independent reading. Researchers need to incorporate orthography into their theories explaining vocabulary acquisition, specifically phonological working memory theories, and they need to attend to its influence in studies they conduct."

Dr. Bill Conrad Apr 22, 2023 04:32 PM

Economically poor children and children of color can read, Tim. The root cause problem is that they are not being taught well. The teaching and administrative cadre are very poorly trained in reading contenr, pedagogy, and assessment skills.

These students get the least qualified novice teachers and the least resources. Racism still abounds!

Much of K-12 marinates in a toxic culture of self over service and loyalty over competence.

The children are generally fine. The adults are messed up!

Time for us educators to look in the mirror and take responsibiliy for student reading failure.

Truth be told.

Enough is enough.

Erin Baker Apr 22, 2023 05:08 PM


What would “an intentional regimen of stretching” look like for 3rd-5th graders? How would you recommend teachers supporting students in these grades as they read complex texts when those complex texts are out of their decoding abilities? What does that instruction look like?

Timothy Shanahan Apr 22, 2023 05:22 PM

One thing you can do if the complex texts are beyond the students' decoding abilities (if they have their basic decoding skills -- like a beginning second grade reading level) is to do fluency work with the texts prior to focusing on comprehension. Studies have found that to be a powerful enabler. Also, guiding students to take on the text in small pieces -- section, pages, even paragraphs.


Timothy Shanahan Apr 22, 2023 05:24 PM

I'm well aware that kids living in poverty can learn to read (I've been dealing with that for more than a half century). Nevertheless, to not acknowledge that it tends to be more difficult to teach these children is unfair to the teachers who strive daily to meet their needs seems insensitive.


Laura Domingo Apr 22, 2023 05:25 PM

I teach grades seven and eight reading and writing. Here is what I will apply from this:
1. My students need to read at higher levels that offer new vocabulary, etc. Push this!
2. My students need to stretch their time engaged in reading and writing on their own.

I can do this! Thanks!

Timothy Shanahan Apr 22, 2023 05:27 PM


I don't know how you could come away with that from this. You might want to read it again.


Timothy Shanahan Apr 22, 2023 05:36 PM


One of the reasons why the NAEP scores are so important is because they continue to link their new tests to the testing that has been done since 1970. Indeed, online testing tends to result in slightly lower scores (probably for the reasons that you mention), but that is then adjusted for. The state assessments are not necessarily as sophisticated in this regard (and that would differ by state, of course). As the states have made sure there were resources to test students electronically, it would be a good idea to make sure there are opportunities to teach students to read on electronic devices (since that does introduce differences). Nevertheless, I'm really not worried about the test performance as the reading performance -- and I believe these tests do reflect that large percentages of our kids are not reading very well when they leave our high schools.


Leslie Laud Apr 22, 2023 05:44 PM

Implementation support is the key. We know what works. Decades of research must guide us. The IES Practice Guides distill it (https://ies.ed.gov/ncee/wwc/practiceguides/). They key is getting this into classrooms. When schools adopt evidence based practices, they see powerful ELA score increases, but they need to ensure the practices are implemented well in all classrooms. Implementation Science offers such hope in helping us do this! And - we need a combination of the most effective kind of phonics instruction (not all are equal) in early grades. As students crack the code, ensure they read complex, rigorous, knowledge-rich texts, and use the most effective strategies (not all are equally effective) for supporting students in understanding and retaining what they read. Teach knowledge because it's important to do, even if knowledge itself isn't likely to raise outcomes. See Cumberland RI for a district that doubled its proficiency (rose from 40 to 80%) after doing this. https://www.schooldigger.com/go/RI/district/00270/search.aspx

Dr. Bill Conrad Apr 22, 2023 05:57 PM

Hi Tim,

I acknowledge your vast experience in education. And that you personally have a belief that economically poor and children of color can read.

Sadly, though we have a tendency to blame the victims rather than improve our practices. We then engage in teaching pathologies where we teach the “poor” children 2 or 3 grade levels lower in order to accommodate their poverty.

Additionally, the system engages in racist practices by assigning their novice and least qualified teachers to classrooms with children of color reserving their most qualified teachers to classrooms with white children. Children of color get the least resources as well. Truth be told.

I have seen this racist pathologies in school districts across the country.

The NAEP has shown us that almost 50 million 4th graders over 20 years are not proficient in reading. The canary is dead in the mine and in fact is in rigor mortis. We need not rage against the thermometers.

Ensuring that our children of color and economically poor children get the most qualified teachers who know their content well, can teach and assess well will ameliorate the pernicious achievement gaps.

The teacher is the key. Let’s focus on that end rather than the “poor” children end.

Richard Innew Apr 22, 2023 09:25 PM

I submit that evidence from some high poverty schools in the Appalachian region of Kentucky show poverty is not really a valid excuse. The problem, which is correctable, is that many teachers don't know how to teach reading properly. Once they are retrained with what science shows works best, the results even in low SES schools can be remarkable. More here: https://bit.ly/3G2pMwA

Helen Apr 22, 2023 11:28 PM

Thank you for this post. I teach third graders and I would like to be better at teaching reading using what you suggested. Can you recommend any books I could read or classes I could take to learn how to implement these things in my reading classes?
Thank you

Kevin Wheldall Apr 23, 2023 03:41 AM

Coming from Australia, I’m puzzled by what is meant in the US by grade level text. How is it defined and how is it determined (and by whom)? Not readability formulae surely …? And can we trust publishers’ leveling?

Mary Hough Apr 23, 2023 05:30 AM

Training in Orton-Gillingham methodology, or in the Wilson Reading/Language program, would help you to teach according to the science of reading. All students can benefit, although more “proficient” readers may not need that supplemental instruction, as it IS supplemental, not to be used as a replacement.

Timothy Shanahan Apr 23, 2023 02:24 PM

Indeed, they do use readability measures. However, as you suspect, historically they have played games with those -- making minor changes to passages to fit the formulas rather than making the texts more readable. In any event, teachers are strongly urged not to teach lower performing students with grade level texts.


Timothy Shanahan Apr 23, 2023 02:26 PM


I would suggest that you start with this website -- you can identify topics that you want to pursue by typing the terms into the search function (magnifying glass). You also can find useful information in the publications and resources sections.

Good luck.


A.J. Saph Apr 24, 2023 04:11 PM

Thank you for another great article. Could you please elaborate on what "specific skills in reading comprehension" you would get rid of?

Gaynor Apr 25, 2023 06:53 AM

What is wrong with readability measures if used properly.? From !970 ,before Marie Clay wrecked things, NZ used a list of 2000 nouns graded by frequency and taken from words found in children' s written work of the various primary (elementary) school ages. The complete list of words taken from the written work was used to compile corresponding leveled spelling lists. These were complimented by materials of suitable difficulty level . This material included home grown NZ stories as well as the age graded US Ginn 100, Scott Foresman and McKee basal readers .These readers were used for 20 years throughout NZ for the lower half of the class from year 3, ( Grade 2 ) on-wards. I think these historical references are interesting in determining whether reading material standards have declined.

Timothy Shanahan Apr 25, 2023 04:16 PM

Readability measures have many worthwhile uses... however, there are two problems with them as far as establishing student reading levels. One, they have never been validated in terms of their ability to place students into texts that provide any kind of learning benefit. They are validated in terms of placing texts on a continuum of reading comprehension. In other words if a formula says 3rd grade or 750L or G that simply tells you people are more likely to be able to comprehend that text well than they would a text with the designation 5th grade, 1000L, or M. It says nothing about the value of those texts for teaching reading to particular kids.

Second, even if they were so validated, they wouldn't be useful for setting kids levels because they are not very precise. Typically there are big confidence intervals around those levels (meaning that 3rd grade isn't significantly different than 2nd grade or 4th grade, and 750L really means 600L-950L, etc.



Timothy Shanahan Apr 25, 2023 05:23 PM


These should give you a clear idea of the supposed "comprehension skills" emphasis that I would drop:



Anita Kim Venegas Apr 26, 2023 01:57 AM

Hey Tim,
I checked the 2022 NAEP for reading. The teacher quoted, only 37 percent of students are reading at grade level. I read the results as 37% are reading below Basic. Is basic different than proficient? And as American fourth gradersm, they've essentially held steady since 1992. Does the NAEP test have a moving target? We're in Colorado, and the change from 2003 CSAP tests to 2012 CMAS tests was a huge learning curve for students.

Timothy Shanahan Apr 27, 2023 02:16 PM


1. NAEP 4th grade scores have not stayed steady since 1992... 1992 was the low point (lower than they had been in the previous 20 years and lower than they have been since). Throughout the 1990s and up to and including 2006, those scores climbed steadily. They then stayed steady until a small drop just before COVID and a bigger drop since. The scores for 8th graders have stayed pretty steady over the past 50 years.
2. 37% is the proportion of kids who are below basic, and I would say those students are clearly below grade level in their performance. The differences in basic, proficient, and advanced are not so easily connected to grade level. Basic has often been described as being the level at which students can do much, but not necessarily all, of their grade level work. The 37% is the scary number because it matches reasonably well to the proportion of young adults (21-25) who are not able to meet their literacy needs (in terms of higher education, work, social participation).
But to tell you the truth, for my own kids and grandkids, I aspire for Proficient or higher for them because I want them to be able to do more than meet their basic needs. I'd also like to see a higher proportion of American kids able to take part in the economic and social benefits of the nation.


rob ackerman Jul 11, 2023 06:38 PM

We need to see actual schedules to understand statements such as, "In my schools, I required 120-180 minutes per day of reading and writing instruction. I know that’s a lot, but it is accomplishable in most schools (and if you get rid of test prep and extended specific skills teaching in reading comprehension and other things that don’t enable students to read better, meeting those times can even be easier).

Timothy Shanahan Jul 13, 2023 05:56 PM

I can't get that kind of information into every blog, but I would suggest that you search my site for more information on classroom organization, Chicago Reading Framework, etc. You should be able to find plenty (search the blogs and look in the publications section). Good luck.


What Are your thoughts?

Leave me a comment and I would like to have a discussion with you!

Comment *

Why Aren't American Reading Scores Higher?


One of the world’s premier literacy educators.

He studies reading and writing across all ages and abilities. Feel free to contact him.