CARROLLTON, Miss.— In the weeks leading up to Mississippi's new third-grade literacy test this past April, Zanysha Amos, age 9, was a bundle of anxiety. She couldn't eat; she couldn't sleep. Ever-present worry furrowed her little brow.

When Zanysha sat down in front of the computer screen to answer the 50 questions on the statewide test, she knew that she could be held back if her score wasn't deemed adequate. The thought was too painful to bear: another year in the same grade, a year older than everybody else, going over the same stuff for 10 agonizing months?

Her mother said Zanysha also had a nagging question she needed answered: What was the point of the As and Bs she pushed herself to get on her report cards, of coming home every day after school and unfailingly diving into her homework, if an uninspired performance on a two-hour test could wipe away a year's worth of effort?

"She told me, if she was studying and doing good work during the school year, then she shouldn't have to pass a test to go to the next grade," said her mother, Ztearreyo Amos, an assistant teacher at Zanysha's school in this north central town just off U.S. Highway 82. Amos had a hard time arguing with her daughter's logic.

Zanysha's questions illustrate the worries that many parents and educators in Mississippi have expressed since Gov. Phil Bryant signed legislation two years ago that declared third-graders could not be promoted without passing a test to prove they were adequate readers. Is such a high-stakes trial appropriate for 8- and 9-year-olds, so tenderly lacking in emotional maturity? Aren't they too young to confront two make-or-break hours on a single test that could keep them from moving up a grade with their classmates?

Bryant believes he is doing the children of Mississippi a favor. He reasons students have little chance of academic success if they can't read by fourth grade and predicts such students run a much higher risk of dropping out later on. The argument is personal to the governor, who recalls the many benefits he derived from being held back when he was a struggling reader in third grade.

Yet, in a state that has historically come up dramatically short in adequate funding for schools, popular sentiment among parents isn't with the governor. Per-pupil spending here is always among the lowest in the nation, while students are some of the poorest in the U.S. In recent interviews, more than a dozen Mississippi parents and educators said they believe Bryant's plan is particularly unfair and will punish 9-year-olds for the results of decades of funding shortfalls.

'Just Not Fair'

A system of school funding under which districts in richer areas receive more funding than districts in poor areas compounds the problem. Karen King, whose son Trey is a third-grader at Marshall Elementary School in Carrollton, said she is stunned by how many more resources children have at a better-funded school about 30 miles down the road, where she teaches sixth grade.

"Those kids have access to so much more," she said. "These kids here at Marshall don't even all have books in the classroom. The difference between what Marshall has and what they have down the road is completely unrealistic. And it's just not fair."

In 1997 the state Legislature established a formula—the Mississippi Adequate Education Program (MAEP)—for determining how much money is required to adequately fund public schools. In the past 18 years, Mississippi's public schools have only been adequately funded twice. The shortfall since 2009 has totaled more than $1.5 billion.

Bryant's initiative comes at a time when state voters will have a chance to demand better financing of the state's school system via a constitutional amendment. Voters in November will be asked to choose between two competing ballot initiatives. Initiative 42 mandates that the state Legislature increase funding for Mississippi schools over the next seven years to reach the level deemed adequate, with the courts enforcing the mandate. The Legislature added an additional initiative, 42A, to the ballot. It will ask voters to decide if the Legislature should establish and support effective public schools, "but not provide a mechanism to enforce that right." Supporters of Initiative 42 say the Legislature added 42A in an attempt to confuse voters.

Many parents believe better-funded schools will lead to better test scores. The Parents' Campaign Research and Education Fund collected numbers that show statewide reading scores for fourth-graders rose in 2008, a year after the Legislature fully funded schools in 2007.

Kids Get Bad News

On May 7, the Mississippi Board of Education decided what the passing score would be for the reading test third-grade students took in April, and schools began the delicate task of delivering bad news to thousands of parents and students.

Statewide, 14.83 percent of third graders—5,612 students—failed to reach the minimum score needed for entrance into fourth grade on the first test. Those 5,612 students took the test again May 18 to 22, and the computer notified them immediately whether they passed or not, said Patrice Guilfoyle, communications director for the Mississippi Department of Education.

Those who failed a second time will be scheduled to attend summer school starting in June and will get a final chance to pass the test in late June or early July. If they fail a third time, they'll have to repeat third grade.

Some districts said they would also invite students who barely passed the test to attend summer school, so they will be better positioned to do well in fourth grade. The state Department of Education stressed that the cutoff score on the test, referred to as the 3rd Grade Reading Summative Assessment, indicates the point at which a student has acquired the minimum reading skills needed to learn properly in fourth grade.

When the scores were first delivered, the failure rate—as everyone expected—was much higher in high-poverty districts, where many children start school behind and stay behind.

In Greenville, a district abutting the Mississippi River in the heart of the Delta, in which nearly all the students at the elementary schools qualify for free or reduced-price lunch, the overall failure rate was 37 percent—with several schools showing rates over 40 percent and one school, Webb Elementary, over 50 percent.

Goodman Pickens Elementary School in Central Mississippi's Holmes County saw a whopping failure rate of 59 percent, while two other Holmes County schools had failure rates of 46 and 52 percent. Holmes County has a median income of $22,325 and the lowest life expectancy of any county in the United States, largely due to its abnormally high obesity rates.

Jackson—the state's largest city with 30,000 public-school students and 34 elementary schools—had a failure rate of 28 percent. And at Carroll County's Marshall Elementary, where Zanysha attends school and where more than 90 percent of the students qualify for free or reduced lunch, the failure rate was 19 percent—far lower than Carroll County officials had anticipated.

The day the scores came out was a good one for Zanysha Amos and her mother, despite their many questions and concerns. Zanysha passed.

The Florida Model

Many education experts look at Mississippi and claim its academic failures are directly connected to a dilemma with which America has long struggled: how to educate the poor. To Gov. Bryant, the answer is to get tough with the little ones and impose what might seem like severe sanctions if they don't measure up. The governor views the get-tough policy as an act of compassion, linking it nostalgically to the extra help and caring instruction he received when he was held back, and noting that holding the line now will help students in the long run.

Bryant's case had a happy ending: His fourth-grade teacher discovered his reading difficulties were a result of dyslexia, and with additional help, the young Bryant was able to overcome his disability.

"Repeating the third grade was the best thing that could have happened to me," the governor said in February 2015, according to media reports—though in 2012 he had recounted that being retained was a "difficult, horrible experience."

Bryant said he modeled Mississippi's third-grade gate after the success of a similar program in Florida, the 2003 brain-child of former Gov. Jeb Bush. A 2012 report by Martin West of the Harvard Graduate School of Education revealed that Florida third-graders who were retained actually outperformed their peers who just barely surpassed the cutoff score over the next few years, though the difference between the groups eventually disappeared in later grades.

But there are at least two key differences between Florida and Mississippi.

Florida pumped about $1 billion into the education system to boost literacy in the early grades, including placing a literacy coach in every school, while Mississippi thus far has added less than $25 million—though Bryant has pointed out that Florida only put in $10 million in the program's first year.

For Mississippi, the money meant less than 100 literacy coaches for the state's 426 elementary schools. Bryant has been defensive when asked why the effort was so small.

"We knew we couldn't put reading coaches in every school," he said. "The question is, 'Why do we put a reading coach into schools? Are the teachers we have now not prepared to teach children to read?'"

Though Bryant seemed to dismiss the necessity of reading coaches, the bill he signed in 2013 mandates intensive reading instruction and support for students who repeat the third grade, including smaller classes and a "high-performing teacher" with a demonstrated record of student improvement in reading.

In the Florida study, author West noted that test-based promotion policies are most likely to be successful "if they are accompanied by specific requirements that retained students be provided with additional, research-based instruction in reading and adequate funding to implement those requirements."

Another key difference between Florida and Mississippi, as West notes, is that Florida allowed an exemption from retention based on a student's portfolio of work—and that in the first six years of the program slightly more than half of the students who failed the test received an exemption.

Mississippi allows exemptions: special-education students, students still learning English and those who have already been retained once.

Yet the state doesn't give educators any discretion in promoting a student based on his or her portfolio of work—totally removing the classroom teacher, who knows the student best, from the equation.

That's an enormous difference in the eyes of students and parents.

Many of Mississippi's poor districts, like Carroll County, struggle to get enough books for every student. Educators wonder where the cash-strapped districts will find the money to hire the additional teachers needed to reduce class sizes, as the governor's legislation mandates.

And where will the districts find these "high performing" teachers, especially if there is no extra money to entice them?

Does Retention Help or Hurt?

For the past three decades, states have been using standardized tests to make a whole lot of extremely important decisions about teachers and students. Which teachers are inadequate? Which teachers are exceptional? Which students should be placed in remedial settings? Which students should be primed for long-term success? The answers generally come from the student's standardized test scores.

There are now 14 states that require third-graders to pass a statewide reading exam to be promoted to fourth grade. But eight of these states accept a portfolio exemption or allow the child's teacher or principal to overrule the results of the test.

West's Florida study notwithstanding, education experts such as Stanford University professor Linda Darling-Hammond, a nationally recognized expert in education policy, claim that retention does more harm than good.

Darling-Hammond said that 30 years of research have told us precisely what will happen in Mississippi: The increased focus on reading in the early grades will lead to a negligible increase in reading test scores in the first few years of the program, but in the long term, the dropout rate will increase as those third-graders who were held back become further disconnected from school.

"We have had dozens and dozens of studies on this topic," Darling-Hammond said. "The findings are about as consistent as any findings are in education research: The use of testing is counterproductive, it does not improve achievement over the long run, but it does dramatically increase dropout rates. Almost every place that has put this kind of policy in place since the 1970s has eventually found it counterproductive and has eliminated the policy. Unfortunately, policy makers often are not aware of the research, and they come along years later and reintroduce the same policies that were done away with previously because of negative consequences and lack of success."

Examples of the studies Darling-Hammond refers to are not hard to find.

For instance, a 2005 report from the Center for Child and Family Policy at Duke University found that grade retention can increase the risk of dropping out by 20 percent to 50 percent.

Supporters of Bryant's approach point to their own studies, which indicate that students are much more likely to drop out if they are unable to read by fourth grade.

Nicole Webb, communications director for the governor, said that Mississippi's Literacy Based Promotion Act was developed "in close coordination with a host of national and in-state education experts and advocates."

If districts really want to meet the needs of third-graders, Darling-Hammond said, they would pass struggling readers along to fourth grade and connect them with a teacher trained in a program like Reading Recovery, which has been reliably shown to improve the skills of youngsters who find reading difficult.

Critics might describe Darling-Hammond's suggestions as "social promotion," protecting students' self-esteem by moving them along to the next grade, even if they aren't academically ready. But Darling-Hammond says the dichotomy is a false one.

"People often present this as if there are only two choices—choice one is hold the kids back, and the other is socially promote them without any additional resources or strategies," Darling-Hammond said. "But the third way, the right response, is one in which you identify the resources they must have and ensure they are getting them immediately. They also should look at whether if you sit them down with a book, can they read? Because a lot of kids perform poorly on multiple-choice standardized tests who actually know the material if you present it in a more authentic way."

Darling-Hammond and others also point out that grade retention is extremely expensive, a fact that cannot be ignored in cash-strapped Mississippi. Holding a child back adds an entire year of instruction for that child to the budget.

"Most of the other responses to help kids who are behind cost significantly less than that," Darling-Hammond added.

Who Gets Held Back?

Disturbingly, the pain that a retention policy inflicts is not shared equally.

"The majority of the kids retained we know will be poor and largely African American, and also new immigrants, because that's happening everywhere else," said Robert Schaeffer, public education director of the National Center for Fair & Open Testing (also known as FairTest), the advocacy group that has long fought against the widespread use of standardized tests. "Schools should do better, but they can't do magic. When you have kids coming to school years behind at the starting line because of poverty, because their parents aren't educated, because they don't speak English, it's ludicrous to expect them to catch up by third grade without lots of extra help."

In Mississippi, many children start school at a disadvantage. Until recently, Mississippi was the only state in the south without a publicly funded pre-K program. The program that exists now is tiny: just 6 percent of 4-year-olds in Mississippi attend a state-funded program.

By comparison, Florida provides pre-K to more than 74 percent of its 4-year-olds. A reported two-thirds of students in Mississippi start kindergarten unprepared.

What Districts Already Knew

Parents and teachers argue that not only does the test punish kids who are already disadvantaged, but it also doesn't provide teachers with any new information about their students.

Billy Joe Ferguson, Carroll County's outspoken longtime superintendent, said when he dug into the scores, he saw that the 13 third-graders from his district's only elementary school (Marshall Elementary, where Zanysha Amos goes to school) who fell below the cutoff score were all students his teachers had long ago identified as struggling readers who would likely have difficulty doing fourth-grade work.

In other words, after all the time, expense and stress of an additional state test, the district didn't find out anything it didn't already know.

So what purpose did the test serve?

"The teachers in that classroom give tests all the time," said Ferguson, 67, who—except for four years—has been Carroll County's superintendent since 1996. "They work with the kids, and they pretty well know that you're going to have some who won't pass to the next grade with or without a test because they couldn't do the work. Maybe they're a little slow, need a little more time, maybe the kid didn't apply himself. Sometimes you can have a split home. You can have so many things happen in a child's life that could impact that one test. Hopefully, with the makeup test, you can filter that out. But to me it's just needlessly spending a lot of money."

Ferguson said his $10 million district budget has been cut every year since 2009, when the recession devastated everyone, setting the district back a total of about $3.5 million, money that was desperately needed to help students reach their potential

"Teacher assistants, literacy coaches—things like that I don't have money for. I would love to have a literacy coach," said Ferguson, who made headlines in 2009 when he decided to retire early and start taking his pension although he continued to run the district. Ferguson's "retirement" saved the district about $100,000 a year for his salary and benefits.

Without Ferguson's move, the district was in danger of being unable to meet its expenses. Ferguson now earns a salary of just $18,000, which he mainly uses for travel and expenses.

In the district he runs, most of the buses are 15 years old, the elementary school's 25-year-old roof could give out at any moment, and aging window units are used to cool the elementary school because it doesn't have central air. The walls are cracking so much in one first-grade classroom that it's possible to see through them to the outside.

In addition, the elementary school of 500 students has no assistant principal, no guidance counselor and only a handful of assistant teachers, who all the early-grade classrooms share.

When schools are strong and well-funded, a child struggling with reading in first grade is more likely to be put on a conveyor built of special services and interventions—reading coaches, reading recovery specialists, tutors, mentors—so that the child is proficient by the time he or she reaches third grade.

At poorly funded schools like Marshall, a child is often left to struggle alone, with very few of these needed interventions. So by the time a test comes at the end of third grade, the student is already far behind.

Rana Mitchell, Ferguson's assistant superintendent, said if the state is going to hold back students who can't read, it should be done in first grade when students are still trying to acquire the skills.

`I Have Book Glue'

To see a picture of scarcity and deprivation, walk into the library at Marshall Elementary, where librarian Ruby Burkhead has set aside a special drawer in her desk for her book-repair materials.

Burkhead said she received just $900 this year to buy new books for the library. Anyone who has spent an hour in Barnes & Noble book shopping with one child knows how painful on the wallet it can be. Nine hundred dollars doesn't come close to getting the job done for 500 children. In a state focusing on early reading skills, Burkhead's library is ground zero—the place where fledgling readers can discover the world of books and become skilled and enthusiastic readers.

Since she can't buy new books, Burkhead has taught herself how to be a book restorer.

"That's what I do on Saturdays—I come in here and repair my books," she said. "The children have to have books to read, so that's what I have to do. I have book glue, I have tape. I can get two or three more years out of them before I have to order some more."

Trey's Bad Day

On a recent spring day, Karen King stood in the main office at Marshall with her shoulders slumped. The school's principal, Fletcher Harges, had just told her that her son Trey had not passed the third-grade test. After all the sleepless nights and upset stomachs, he still got bad news.

"The morning of the test was actually the worst part. He was nauseated, sick, because he had to take the test," King said. "It was terrible, one of the worst feelings I ever felt in my life, because there was nothing I could do about it. It's something he has to do. Mr. Harges just told me he is going to have to retake it. Thinking of him going through all of that again just sucks."

Since Trey is already in special education because of learning disabilities and has already been held back once, he will likely receive an exemption and be promoted to the fourth grade anyway. But the damage has already been done.

"He's overcome a lot of his issues, but what does this do to him long term?" his mother asked. "To his self-esteem, his social development, everything? Even if he does pull through it, I think it has had negative effects on him."

King paused and looked out the window. "There must be a better way to assess children's knowledge than put them through this kind of stress," she said. "It's just absolutely not fair."

This story was produced by The Hechinger Report, a nonprofit, independent news organization focused on inequality and innovation in education.