One of the key features of the “Mississippi Miracle” is the retention of third-graders who do not score well enough to enter fourth grade. Third-graders with low reading scores are held back for an extra year.
Critics of the “Miracle” say that holding back the lowest scoring third-graders inflates the fourth grade scores.
But what about the effects of retention in the students who are held back?
Matt Barnum of Chalkbeat reports on a new study that found negative, long-term effects of third-grade retention.
It’s an age-old debate with an emerging conventional wisdom: Third graders should not move on to the next grade if they are still struggling to read.
There’s both logic and evidence behind this policy. Studies have found that students have higher test scores after they’re held back. This practice may also have played a role in helping Mississippi make remarkable improvements in recent years. A chorus of policymakers and journalists have insisted with growing confidence that others should replicate the state’s model.
But a new study offers a warning about the downside risks of retention. Third graders who had to repeat a grade in Texas were far less likely to graduate from high school or earn a good living as young adults, nearly two decades later. The harmful effects were quite large and came despite initial improvements in test scores.
“Retaining low-achieving students in third grade further deepens educational and income inequalities,” writes Jiee Zhong, an economics professor at Miami University.
The findings are hardly the last word on this topic. But they complicate the evidence base for retention at a time when more states — like Arkansas, Indiana, and West Virginia — are adopting this policy.
The paper, set to be published in an economics journal, examines an early 2000s Texas policy to hold back struggling readers. Students had three chances to pass the state exam.
Zhong, the researcher, looked at those who just barely missed the passing score versus those who just reached it. These students were essentially identical — the only difference was a few questions right or wrong on the test. Yet those handful of questions changed the trajectory of many students’ lives by determining whether they would be held back. This also created a natural experiment that allowed Zhong to compare the two groups of students, thus isolating the effect of retention.
Failing the exam wasn’t a guarantee that students would repeat the grade — parents could seek exemptions — but it dramatically increased their chances. Relative to the overall student population, the retained students were more likely to be low-income, Black or Hispanic, and still learning English.
In the short term, the results were promising. By the time retained students finished fourth grade, their test scores were much higher. But there were warning signs. Students missed more school after they were held back. As the years went on, the test score gains, relative to non-retained students, started to fade. In middle school, the students who had been held back were more likely to exhibit violent behavior (although this remained rare).
By the end of high school, retained students were 9 percentage points less likely to graduate, compared to similar students who weren’t forced to repeat third grade. This is a very large effect. Even those students who graduated typically did so a year later, reflecting the extra year from being held back.
At the age of 26, the previously retained students, now young adults, earned less money than if they hadn’t been held back. Again, the effect was substantial: nearly $3,500, a decline of 19%.
To finish reading the article, please open the link.

The retention debate is old. I had a colleague, now gone, who swore back in the 1980s that our problems where the students were concerned related to their knowing that, no matter what they did, they would be passed on to the next grade. At that time, I was of the opinion that the problem was grades themselves. I argued for grouping based on student involvement: if a student was intellectually mature enough to directly engage in reading Dickens or performing applications of complex mathematical concepts, that should drive our groups of learners.
I have repented of my idealized view of the system. The social aspects of what school does makes my idea more problematic. My colleague never repented of his more Calvinistic approach. He was wrong, of course. Some of these things most intractable problems we faced were sprouted from the great garden of rural poverty, a little known adjunct to “the inner city” and other myths we conjure to justify not doing anything to create economic justice in society. Others were of our own structural making. Most could be laid at the foot of a state government not interested in spending the hard-earned dollars of people with money to burn.
The question of third grade retention is a problem related to misuse of a test. Retention used to be a function of a teacher talking to a parent and determining whether it was a good choice for the child. It was always a conversation about failure that came too late. But it was not based on a flawed test.
LikeLike
I remember how sensitive I was about academic comparisons with my peers. In middle school, my algebra teacher recommended that I take Intro to Algebra instead of Algebra 1, and I was deeply embarrassed. My test scores were solid, but sometimes I didn’t complete homework, which pulled my average down.
Determined to catch up, I enrolled in geometry during summer school. There, I found myself alongside some highly motivated, academically focused students who were trying to accelerate even further — and I had no trouble keeping up with them.
I can’t imagine how I would have felt if I was retained. I’m sure that many students just check out and drop out asap.
LikeLike
If these states bothered to look at the research, they would understand that retention is damaging to students’ self esteem, and it has harmful consequences that may result in students dropping out or becoming discipline problems.
LikeLike
Was the outcome looked at for earlier cohorts that did not contain retained students? The question isn’t what the long term outcome of students retained is, but what the “gap”-∆ among the retained group and their non-retained counterparts is compared with who would have been retained and were not – that gap (that would have been) in the earlier cohort.
And then there’s all the trouble with test-continuity between years in distinguishing this subgroup that differed by “just one-or-two questions”. Perhaps 1-2 questions is significant, but the issue is the treatment; OTOH that test-variability is itself variable (some years those “1-2 q’s” might matter intellectually, some years not).
Yes, transcendental, forever-questions. Also totally need to examine the effect of teachers’ attitudes among the resulting schooling on retainees. etc. Were they placed in a special remedial subgroup? Offered catch-up opportunities? Isolated or segregated in what ways, could have been beneficial, could have been stigmatizing. Could simply be parsing folks with special needs and no resulting intervention. That doesn’t speak to pedagogy…. (well, or maybe it does).
Point is, that outcome may be real but it’s really not clear whether it’s attributed to retention, measurement artifacts, or ancillary treatment differences.
LikeLike
But why 3rd Grade retention instead of another grade? Might this be an attempt to inflate future 4th Grade NAEP scores? (The answer seems rather obvious…) Then get ready to hear all the bragging about how they got education “right” if/when those inflated 4th Grade NAEP scores come to fruition.
LikeLike
D’oh! Just read the Paul Thomas post below this one. Sorry to state what’s already been said.
LikeLike