John Thompson, writing from Oklahoma, reviews the research on the “Mississippi Miracle,” the strategies that boosted fourth-grade test scores? Those strategies are now being turned into law in states across the country. John takes another look.
He writes:
When the so-called “Mississippi Miracle” and the “Science of Reading,” became the drivers of the Oklahoma Strong Readers Act, I pointed out the large body of research-based evidence that its advocates ignored. Yes, there was evidence that the so-called “Marathon” could raise 4th grade test scores, at least in the short-term, but there was little or no evidence that it would do more good than harm. Negative results were even more likely in Oklahoma where our rightwing politics and our minimal amount of funding would drive the experiment.
I also asked experts and journalists about policies that common sense would say violate the norms of social science research. For instance, NAEP tests have been designed to minimize the use of test prep to jack up scores. And, yes, I was told that there are people looking into that issue.
Being a former academic historian, I would read the way that reformers like Adam Tyner and Rachel Canter track NAEP scores. However, I was dismayed by their interpretations of those patterns. After all, whether laws were passed in 2002 or 2013, those mandates didn’t take effect immediately. If we want to learn what policies worked in what places, we need to identify what was actually happening in schools before we can figure out whether those approaches worked.
And that leads me to “NEPC Review: Inside the Mississippi Marathon” by, Gregory Camilli, Professor Emeritus from Rutgers University. Camilli was extremely precise and diplomatic, when reviewing the Progressive Policy Institute’s, “Inside the Mississippi Marathon,” written by Rachel Canter. He explained that it presents an “insider account of how the K–3 education system in Mississippi evolved over roughly two decades.” But, he “focuses on just a slice of the report: its description of the state’s performance in reading on the National Assessment of Educational Progress (NAEP), and, to a lesser degree, the policy implementation.”
Camilli reports that Mississippi’s Literacy-Based Promotion Act (LBPA) was passed in 2013, but “full implementation of the various pillars of the LBPA arguably took place in 2017-2019.”
Then he notes that “the PPI report also presents no research literature on [the] effectiveness of the LBPA,” and “it is not presented or intended as a research study.”
The LBPA prioritized the lowest-performing schools. By 2015-16, full literacy support was given to 18% of approximately 700 elementary schools.
These “designated schools,” were assigned a literacy coach (hired, trained, and monitored by the state)” who modeled and observed instruction. They also received a “reading curriculum that the state had reviewed and approved as consistent with the ‘science of reading.’”
It was estimated that the time available for administrative duty and travel, and coaching, “may equate to approximately 15-35 minutes per week per teacher.”
So, how could Mississippi’s full LBPA gains provide more than a small contribution to NAEP test score increases?
Camilli concludes LBPA “most likely had modest impact on NAEP scores due to the relatively small number of students in designated schools and the even smaller number selected for the NAEP sample.”
Conversely, low-performing students in 82% schools were subject to some assessments, and retention, but not the comprehensive supports that were said to be crucial for the Miracle. Camilli explains, “In non-designated schools, LBPA exposure consisted of a modest state-wide bundle of supports including retention screening.”
So, was it the threat of retention that played the largest role in increasing test scores? Or, was it test-prep that improved scores? If so, is there any reason to believe those scores meant reading levels improved? After all, even though Camilli focused on the 4th grade NAEP scores that “Inside the Mississippi Marathon” emphasized, he noted Mississippi’s failure to improve 8th grade NAEP reading scores.
That leads to the question as to whether reading comprehension improved.
Camilli wrote that threat of retention was a prevalent incentive:
It could be argued that the threat of retention was the primary treatment for the majority of Mississippi students, because students in non-designated schools received only distal instructional support in the form of teacher training. This threat itself may have caused principals, teachers and students to devote more instructional time and effort to literacy.
(Distal instructional support means interventions, guidance, or systemic resources delivered remotely, indirectly, or at an organizational level rather than through immediate, face-to-face coaching.)
And, “This threat (of retention) itself may have caused principals, teachers and students to devote more instructional time and effort to literacy.”
Moreover, “the National Assessment Governing Board (NAGB) suggests that in 2015, the MDE revised its state assessment to align more closely with the NAEP assessment.”
Camilli noted the “vast literature on the outcomes and efficacy of test preparation, which is a notoriously controversial topic,” and the way that “it may lead to increases in test scores without enlarging students’ capabilities.”
And that leads to the question as to what was the effect of Mississippi’s use of NAEP materials in order increase its State Proficiency scores. When defending their failure to increase 8thgrade NAEP reading scores, advocates for the Miracle argue that their scores decreased less than other states. But, it’s possible (or likely) that those Mississippi scores have only declined by 2 points since NCLB became law in 2002 due, in part, to the use of NAEP materials.
On the other hand, from 2019 to 2024, the dubious Mississippi State 8th grade Reading Proficiency rates increased by 6 points, while its NAEP Proficiency scores dropped by 2 points, increasing its State/NAEP Proficiency Gap to 19 points. And the 4th grade gap reached 25 points!
Those gaps again call into the question about whether Mississippi’s data-driven accountability and mass retentions actually improved reading, as opposed to producing short-lived surges in 4th grade test scores. And it further calls into question Mississippi Gov. Tate Reeve’s plan to ramp up those policies up until 8th grade. And I would further ask why Oklahoma should commit to Mississippi’s model, before they can present solid evidence that it has worked.
New reporting and scholarly research, like Dr. Camilli’s, indicate that the so-called “Miracle” is doing even more harm than anticipated. To defend public education, and the joy of reading, we need to better explain to its advocates why the benefits that once seemed possible, and the predictable harms of the “Marathon,” are imposing even greater threats on school cultures. We must also increase our conversations with Democrats and adult Republicans, who have questions about mass retentions of students.
In my experience with education advocates, which I suspect is similiar to education supporters in many Red states, they knew that legislation based on Mississippi’s experiment was bound to become law. So, they focused (usually unsuccessfully) on providing more social supports, like high-dosage tutoring, and better funding in order to provide human contacts, where students would be seen as persons, not test scores.
But, now we must focus on the damage that we now know is being inflicted on K-12 education in Mississippi, and push back against those policies in our states.

As usual, the reforms are only surface. This is because attention to education at all is a state level performative act designed more for getting a certain group elected than helping kids. Because of this, testing becomes the engine that drives reforms. The whole thing is a huge reality show, the most oxymoronic pair of words in our modern language. Like Trump’s famous line (you’re fired) none of this means what it says.
LikeLike
I have been waiting for an unbiased evaluation of the “Mississippi Miracle.” It did seem to good to be true. I thought that any improvement in reading has to be attributed to increased teacher training. Linda Darling-Hammond has pointed out that the best school systems have highly trained teachers who are given autonomy and resources to meet high expectations. Logically it would the teacher training that Mississippi did should have more weight than retentions in “Miracle.” We not accept what Doug Carnine is advocating which is just use and follow what is in a box and everyone will be above average. He does advocate for teacher training but I would not restrict the knowledge teachers need to just the sciences or reading and learning as they are now. Teachers should have a wide array of knowledge and we should also look at Michael Commons’ Model Hierarchy of Complexity to have teachers who are at the systematic level of problem solving.
LikeLike
Read the review. This blog post does not adequately represent my views.
LikeLike
I would encourage readers to read for themselves the full 6,000 word review. It’s not possible to fully grasp context and analysis from short quotes and paraphrases. With regard to test prep, for example, I wrote “I will not attempt to sketch the vast literature on the outcomes and efficacy of test preparation, which is a notoriously controversial topic, other than to say that it varies highly in quality and effectiveness. In the right circumstances, testing—including summative, formative and interim, and classroom—is an essential instructional support. In other circumstances it may lead to increases in test scores without enlarging students’ capabilities—their freedom to do and be things they otherwise could not.”
LikeLike
Gregory, it seems that core to your critique is this: “It could be argued that the ‘threat of retention’ was the primary treatment for the majority of Mississippi students, because students in non-designated schools received only distal instructional support in the form of teacher training… Relatedly, recall that the intervention did not include supplemental contact with students (e.g., tutoring). In short, the mechanism by which most of the students received instruction in reading is not sufficient to explain improved performance on NAEP, and neither is the selection effect.
I don’t see your source for such assertions, the paragraph contains no footnotes.
When I ask Gemini: “Did students in non-designated schools receive any additional tutoring supports after LBPA?” It responds as follows (can you, Gregory, identify any hallucinations here?):
Yes, students in non-designated (non-targeted) schools did receive additional tutoring and intervention supports under Mississippi’s Literacy-Based Promotion Act (LBPA).
While the state prioritized allocating state-funded literacy coaches to the lowest-performing schools (the bottom 20%), the statutory mandates of the LBPA applied to all K–3 public school students across every district in Mississippi. Key Protections & Supports for Students in Non-Designated Schools
Mandatory Screening and Early Identification
Every public school district in Mississippi is required to administer an approved universal reading screener (e.g., i-Ready, STAR, or NWEA MAP) to all K–3 students three times a year.
Regardless of whether a school is designated as low-performing, any student who shows a reading deficiency on the screener must immediately receive extra support.
Required Individual Reading Plans (IRPs)
Non-designated schools are legally mandated to create an Individual Reading Plan (IRP) for any student flagged with a reading deficit.
The IRP prescribes specific targeted interventions, which often include small-group tutoring, high-dosage instruction, or after-school literacy remediation.
Multi-Tiered System of Supports (MTSS) & Intensive Interventions
Under the LBPA, districts must use MTSS to deliver Tier II (small-group) and Tier III (intensive, individualized) interventions.
For students struggling with reading in non-designated schools, districts fulfill these requirements through after-school tutoring, summer reading camps, and dedicated small-group intervention time during the school day.
Mandatory Summer Reading Camps
Districts statewide must offer free Summer Reading Camps to all 3rd graders who do not pass the initial 3rd Grade Reading Summative Assessment.
These camps provide intensive, small-group tutoring before students take the re-test to qualify for 4th-grade promotion.
Universal Teacher Training
Even in schools that did not receive state-funded coaches, teachers were required to complete the LETRS (Language Essentials for Teachers of Reading and Spelling) professional development program. As a result, Tier I classroom instruction and in-class small-group tutoring improved across all schools statewide.
Non-Profit Supplemental Programs
Beyond district-administered interventions, statewide non-profit programs stepped in to expand tutoring access. For instance, programs like Reading Corps mobilized trained community tutors to provide high-impact, 1-on-1 reading support to K–3 students in both urban and rural school districts across the state, expanding tutoring access outside of state-designated intervention sites.
LikeLike
That’s not correct. Please read the article for my thoughts.
LikeLike
Gregory: “That’s not correct. Please read the article for my thoughts.”
What is incorrect? I had read the article. It claims that factors “such as coaching, instruction, and teacher training” “most likely had modest impact on NAEP scores” largely because a “relatively small number of students were in designated schools,” “instructional support for teachers was arguably modest in LBPA-designated schools and less so in non-designated schools,” and “In non-designated schools, LBPA exposure consisted of a modest state-wide bundle of supports.”
However, I did not, for example, find substantiated evidence that Individualized Reading Plans for students in non-designated schools were minimal in their aims, implementation, or results.
You suggest that preparing students to perform well on tests of their reading ability may have been a primary factor in their performing well.
Without effectively distinguishing that, in educational substance or effect, from the efforts of educators aiming to help their students benefit from burgeoning reading skills.
It is legitimate to speculate about such matters, and I would be curious to hear from those closely familiar with NAEP reading tests if they think that scores might be very substantially improved without enhancement of truly useful skills. And if there is any evidence that that happened in Mississippi.
LikeLike
Gregory: “That’s not correct.”
You are right if you are alluding to Gemini overstating the mandatory nature or some of those interventions, like reading camp. Perhaps it extrapolated improperly from material like this regarding a Mississippi school district designated for extra interventions:
“Mandatory summer camps and additional time spent on reading skills in the computer lab are two approaches that SCCSD literacy personnel credit with improving outcomes for its early readers. Prior to entering third grade, all SCCSD students attend a mandatory reading summer camp for one week. Students who do not pass the weeklong summer camp return for an intensive four-week camp prior to third grade. During the school year, elementary students have either a 90-minute or 120-minute uninterrupted block for reading instruction every day where all five components of reading are taught.” https://excelined.org/wp-content/uploads/2019/03/ExcelinEd.MSGatewaytoSuccess.March2019.pdf
But summer reading camps aren’t mandated state-wide. According to item 73 in the State’s LBPA FAQ, “While the law allows for summer reading camps, schools are not required to use this approach” https://www.mdek12.org/sites/default/files/Offices/MDE/OA/OSA/LPBA/lbpa_faq_new_format_final.pdf
While Gemini overstated the breadth of mandated interventions, did you understate them with material like this: “because students in non-designated schools received only distal instructional support in the form of teacher training… Relatedly, recall that the intervention did not include supplemental contact with students (e.g., tutoring).”
Am I correct that you indicated that preparing students to perform well on tests of their reading ability may have been a primary factor in their performing well?
Without effectively distinguishing that, in educational substance or effect, from the efforts of educators aiming to help their students benefit from burgeoning reading skills?
It is legitimate to speculate about such matters, and I would be curious to hear from those closely familiar with NAEP reading tests if they think that scores there might be very substantially improved via test preparation without concurrent enhancement of truly useful skills. How? And is there is any well-documented evidence that that happened in Mississippi?
LikeLike
Dr. Camilli,
Your review of Ms. Cantor’s paper was about 6 times longer than my post. As we discussed by email, my focus was on the evidence you presented on Mississippi’s student supports, retentions, test prep, and use of NAEP materials.
If I was writing a 6000 word piece, I would have quoted the entire paragraph you wrote on test prep. But when I sent you my draft, you had no problem with my statement:
Camilli noted the “vast literature on the outcomes and efficacy of test preparation, which is a notoriously controversial topic,” and the way that “it may lead to increases in test scores without enlarging students’ capabilities.”
LikeLike
The clause omitted from the text was just one example of how compression can substantially alter meaning. Having been away from the reading wars for some time, I appear to have lost my knack for defensive writing, i.e., zealously guarding quotable sentences, or being cognizant that the distinction between words like prevalent and primary may not be apparent to readers. Or that original text will survive in a contest for attention.
LikeLike
Gregory,
Please post a link to the article.
LikeLike
Here is s link to the review.
https://nepc.colorado.edu/review/mississippi
LikeLike
Dr. Camilli,
During our discussion, I realized that the “distinction between words like prevalent and primary may not be apparent to readers” but I followed your recomendation regarding the use of those two words. I don’t see how my following your advice regarding those two words “substantially altered” my use of what you wrote. But if I did, I apologize.
Yes, I got the feeling you didn’t want to address the issues that I wrote about. But, I was communicating with you so I could accurately use some of your findings in my post about the reading wars.
I’m not an expert on data the way you are. But I am a consumer of data-driven research. As I indicated, I focus on combining data-driven and social and cognitive science, as well as my experience, along with other educators researchers’ experience, with the tragic effects of NCLB and other corporate reforms. So, I won’t apologize for compressing the paragraph you cite, regarding Cantor’s claims.
LikeLike
A couple of observations.
First, the claims that Mississippi’s improvement in Grade 4 NAEP Reading have not shown in Grade 8 NAEP are out of date. Interested readers can see a set of tables of NAEP Grade 8 Reading results that show, at least compared to other states, Mississippi has started to move upward in Grade 8 NAEP reading, as well. https://x.com/Innes434/status/2084399875827830818
Of particular interest in these NAEP tables is the data on how many states statistically significantly outscored Mississippi for Grade 8 NAEP Reading in 2013 versus 2024. For white students, back in 2013, whites in 43 other states statistically significantly outscored those in Mississippi. By 2024, whites in only 7 states could still make that claim. For Black students, back in 2013 out of the 42 states that got Black scores reported, those in 27 states outscored Mississippi’s. By 2024, only Black students in 2 states could make that claim.
The other point is that other states that have adopted Mississippi-like reforms are starting to show progress, as well. Yesterday, the Kentucky Department of Education made a presentation to the legislature’s Interim Joint Education Committee on Education, and the presenters happily reported on interesting progress thanks to things like LETRS training for teachers and other programs. You can watch the presentation for yourself on YouTube at: https://www.youtube.com/live/xDlNMqZCTos.
LikeLike
Nobody can say for sure if the Mississippi miracle is due to their interventions or if the results are simply a some type of a “surge.” Unlike scientific testing in which lab conditions are controlled, educational testing on humans is far less scientific. Consider a study I read by Stephen Krashen, where he maintains that self-selected silent reading is as effective as skills-based reading instruction. It is possible that the Mississippi students that were good decoders learned to enjoy reading and started reading more by borrowing books from the public library. Students that struggle with decoding are unlikely to read for pleasure. Perhaps some of the students were reading for pleasure when they were in 8th grade, and that is what accounts for some of the surge. Who knows for sure how much of the results on tests were the result of the school efforts or what students did for themselves?
Likewise, we should dispel the myth that the so-called science of reading is the only way to teach phonics. As a reading teacher, I have taught many students to understand phonics using other approaches that were extremely successful. Once students can decode with automaticity, they will start to gain confidence in their abilities, and they will start to enjoy reading which is more likely to result in them becoming life long readers.
LikeLiked by 1 person