The Education Gadfly Weekly

Many moons ago, one of us was a first-year teacher of high school English in a rural district that loved its football. It was the end of the first term as I sat at my desk after school in a musty classroom trailer—there being no room in the main building for rookies—trying to finish up my grades in time for report cards. I heard the steps creak and looked up to find Mr. Simpson, the brawny football coach who’d been at the school since forever, interrupting my calculator punching (this was the Mesozoic era before online gradebooks). Honestly, my first thought was one of alarm, as I was pretty sure Coach Simpson hadn’t said one word to me all year. Yet here he was after hours in my low-rent trailer with no one else around.

Thankfully, he was feeling just as awkward as I was, so he got straight to the point: “I understand that Darrell is just a couple points shy of getting a C this quarter in your class. I know we don’t want to give him a grade that he doesn’t deserve, but I was just wondering whether you had any make-up work for him to do in study hall, or maybe he could re-take that last test again after he studies some more.” My perplexed expression must have signaled that I was no whiz in reading between the lines, so he continued: “I’d personally make sure that he does every assignment in study hall and gives it his best shot. I hate to ask and all, but he’s one of our best players, and we’re really going to need him for the playoff game next Friday.”

So there it was. You see, our high school back then had a policy that athletes had to keep a C average in every class or couldn’t play sports—at least not until they got their grades back up.

I wish I could say that I promptly told Mr. Simpson to take a hike. After all, Darrell had been slacking all fall, barely getting by, though he was clearly able to do much more than execute a perfect pass. I should have responded, “I’ll let both Darrell and myself down if I communicate to this young man that I’m willing to accept less from him than I know he’s capable of!”

Sadly, I didn’t say any of that. I was a twenty-something newbie trying to keep my head above water in a sink-or-swim school. So I dutifully agreed to pull together Darrell’s study-hall packet and breathed a sigh of relief as Coach Simpson shut the door behind him.

We tell you this true tale because it underscores the motivation for our newest study and the complexity surrounding the issues it delves into: high school grading practices and how they intersect with teachers’ expectations for their students, and the impact they have on student outcomes.

The limited prior research on this topic shows that instructors who recognize and believe in their students’ potential—and maintain high expectations for them—significantly boost the odds that their students will go on to complete high school and college. That’s what American University’s Seth Gershenson and his colleagues found in a previous study that used teacher survey data to define expectations.

Another way to define expectations is to measure how teachers approach grading—specifically, whether they subject students to more or less rigorous grading practices. A lone study conducted sixteen years ago by David Figlio and Maurice Lucas in one Florida county found positive academic and behavioral impacts for nearly all students from elementary teachers’ high grading standards. Still, given the central role of grading in U.S. schools, we know shockingly little about how it impacts a child’s future, and virtually nothing about grading standards in middle and high school.

Dr. Gershenson’s existing work on teacher expectations, as well as his prior study for us on grade inflation in high school, made him an ideal partner to tackle this neglected area of research. Like him, we were interested in whether a teacher’s approach to grading students’ work affected their outcomes in the short and long terms, and whether those standards differed by teacher, student, and school characteristics.

Gershenson addresses these questions in Fordham’s new study, Great Expectations: The Impact of Rigorous Grading Practices on Student Achievement. Specifically, he investigated the following: How do the grading standards of an Algebra I teacher affect students’ content mastery, as measured by their performance on the end-of-course exam? Do the grading standards of an Algebra I teacher impact students’ performance in subsequent math courses like geometry and Algebra II and their likelihood of graduating from high school? Does the impact of an Algebra I teacher’s grading standards vary by pupil, school, or teacher characteristics? And what school and teacher characteristics predict teachers’ grading standards?

His data come from the grading standards of eighth and ninth grade Algebra I math teachers in North Carolina public schools. Algebra I is ideal for this purpose, as it was a state graduation requirement for the eleven-year period that the study covers (2006–2016), and it also had an end-of-course (EOC) test during those years. Having both course grades and EOC scores allowed Gershenson to define teachers’ grading standards in a straightforward manner: Teachers who inflate grades—meaning they assign good grades to students who perform relatively poorly on the EOC—exhibit low standards, while teachers who assign lower grades than we might expect given students’ exam scores exhibit high standards. He compared students of teachers with higher grading standards to their peers who have teachers with lower grading standards but still take the same course (Algebra I) in the same school, in the same grade, and in the same year.

There are five key findings:

1. Students learn more from teachers who have higher grading standards.


2. Teachers with higher grading standards improve their students’ performance in subsequent math classes up to two years later.


3. The positive effect of higher standards impacts students of all types and in all kinds of schools. Whether black, Hispanic, white, male, or female, students learn more when taught by teachers with higher grading standards.


4. Teachers who attended selective undergraduate institutions, hold graduate degrees, and have more experience tend to have higher grading standards.


5. Grading standards tend to be higher in suburban schools, middle schools, and schools serving more advantaged students.

The study includes Dr. Gershenson’s take on what these findings mean for policy and practice—a couple of which we underscore and extend here.

First, we should use information about grading practices to improve instruction. Teachers are not to be blamed for having low grading standards when many of them don’t know where to set the bar for high-quality student work. This is not a major focus of most teacher preparation or professional development. Education Trust has provided a valuable service by taking a closer look at what teachers assign students and asking whether those tasks reflect today’s higher academic standards. Similar questions should be raised about teacher grading practices.

Likewise, schools and districts would do well to share with teachers how their grading standards compare to the standards of other instructors teaching the same subjects and at the same grade levels. Teachers need to know whether their expectations fail to match—or possibly surpass—those of their colleagues. Educators might be more willing to aim higher if they knew they were off target and taught how to get closer to the bullseye.

Second, let’s not forget that none of this is possible without an external measure of student performance. The simple definition of grading standards used in this report can be easily calculated by schools and districts—but only if they have a summative test, such as an end-of-course exam. The current angst about over-testing has likely resulted in the recent dip in such tests administered across the country.

Let’s be honest. Most of us want teachers to have high expectations when it comes to grades, but we’re gradually making it harder, not easier, for teachers to do that. Case in point: More than one thousand colleges and universities have adopted test-optional admissions policies, arguing that college entrance exams provide an unfair advantage to middle- and high-income students. Not requiring them, they say, expands access for poor students and students of color.

There’s plenty of debate about whether that’s true, but one thing seems clear: Grade-point averages will now matter even more, so it’s crucial that they be accurate representations of a student’s academic performance. The current push for test-optional college admissions makes it that much more difficult for high school teachers, who now face even greater pressure to be easy graders to help their students get into selective colleges.

That’s a big problem. Think about it: If there’s pressure on teachers from one of their own to inflate a grade for a kid to play in next week’s football game, just imagine what that pressure looks like from his parents to get him into a good college.

Florida’s Tax Credit Scholarship program has provided more than 780,000 scholarships since its inception in 2001. The largest program of its kind in the nation, it allows dollar-for-dollar tax credits to companies that contribute their eligible tax dollars to fund scholarships, which pay for children from low-income families to attend one of approximately 1800 independent and religious K–12 schools in the state.

Last month, the Orlando Sentinel reported that 156 private religious schools participating in Florida’s program “have policies that say gay and transgender students can be denied enrollment or expelled or that explain the school opposes their sexual orientation or gender identity on religious grounds.” Program officials insist that count includes schools affiliated with churches or organizations that simply have a statement of faith about marriage or homosexuality, and that the actual number is far smaller. Furthermore, there have been no claims of LGBTQ scholarship students denied admission to participating schools. Regardless, after being contacted for comment by the paper’s reporters, companies including Wells Fargo, Fifth Third Bank, GEICO, and Waste Management announced they were either withdrawing support or called for a state review of the scholarship program’s eligibility rules.

Advocates are putting a brave face on the fallout, noting that many of the companies that have withdrawn their support are either minor donors or haven’t contributed in years. But the damage is real enough and significant. Florida’s Tax Credit Scholarship program serves 108,570 economically disadvantaged students, most of whom are black or Hispanic. “Every $1 million lost because of this attack on choice is 140 fewer scholarships for low income kids, most of whom are students of color,” notes Patrick Gibbons of Step Up For Students, the non-profit organization that administers the program.

Privately, program and parent advocates are worried that the fallout could be dire. They’re wise to be concerned. Big companies are sensitive to adverse publicity. The last thing most want is to be caught in the crossfire between advocates charging anti-LGBTQ bigotry, and those who see anti-religious bias among scholarship opponents. Fifth Third Bank, which contributed $5.4 million dollars in 2018, “quit donating after [state representative] Carlos Smith got fifteen retweets,” one choice advocate told me. “If donors will bail on low-income minority students that easily, then people should be afraid.”

What’s happening in Florida portends a new and divisive raising of the stakes in the battle over school choice that could echo beyond Florida. Having lost in the legislature and in the courts, a new playbook is being written before our eyes: Make ESAs too hot for donors to touch. Choice proponents can argue that scholarship funds are private dollars. They have the facts and the law on their side. But if donations are weaponized, it might not matter to image-conscious donors.

A perverse consequence of this battle is that students who actually have faced discrimination and bullying may be denied a safe haven, to guard against the theoretical and unlikely chance that some future student might want to enroll in a school where he or she might feel unwelcome. Ron Matus of Step Up for Students recently described the case of Elijah Robinson, a mixed-race, openly LGBTQ teenager who was bullied and harassed at his public school before transferring to a private religious school with tax credit scholarship funds. “If I had stayed at my previous school,” Robinson told Matus, “I honestly think I would have lost my life.” That story prompted defenders of Florida’s scholarships to take to Twitter to accuse lawmakers opposed to the scholarship programs of supporting LGBTQ students in theory but denying them protection in practice.

The challenge for serious people and policymakers is to look beyond the partisan passions of the moment and movement orthodoxies on both sides and ask what is likely to increase tolerance and lead to a fairer and more equitable society in the long run. The available evidence tilts in favor of educational pluralism and school choice as the more potent mechanism. Because we are accustomed to the default setting of government-funded and government-run schools, mostly zoned by geography, it is natural and inevitable that Americans would be suspicious of private and religious education, and to assume it contributes to insular self-sorting, intolerance, and anti-civic behavior. Just the opposite appears to be true.

A persuasive body of research links private schools, particularly Catholic schools, to superior civic outcomes, including greater tolerance. Arguing for the long view may be uphill work in our era of scorched-earth, base-pleasing politics. However, as Ashley Berner of Johns Hopkins observed in her 2017 book, No One Way to School, pluralism “encourages distinctive school cultures, the sine qua non of strong academic and civic outcomes.” If we take seriously the moral foundation of our children, the evidence favors pluralism. When children receive the same messages of tolerance from multiple authorities—at home, school, and church—that foundation is more likely to be built and to stick.

Does this mean every religious school is better at breeding tolerance than any public one? Obviously not. But let’s accept as sincere those who object to Florida’s tax credit scholarship program and similar vehicles on grounds that they might discriminate against LGBTQ youth. If the object is to protect vulnerable kids, and not merely to defend at all cost traditional public schools, it serves those kids’ long-term interest to support—or at least not aggressively oppose—a small number of schools some might find objectionable and to take the long view.

Interesting question.

Before I answer, let me ask one: What keeps Jeff Bezos, the founder of Amazon, up at night?

You know Amazon, the trillion-dollar corporation that delivers something like a five billion packages a year.

I’m at a professional meeting. The chair asks what “levers” we have for improving reading achievement in the U.S.

It’s an easy question. There are so many possibilities.

The first one most folks think of is the teacher. If teachers did better, kids would do better.

There are a lot of alternative levers: school administrators, politicians, bureaucrats, publishers, universities, assessments, standards, curricula, media, screens, mom and dad.

As these discussions go, this one isn’t bad. Lots of levers, little blame.

But I’m not sure the levers question is the right one. I’ve grappled with all of those levers—“successfully” sometimes.

And, yet, as relevant as each and every one of those can be, I’m thinking about what Jeff Bezos worries about.

What’s that?

The “last mile problem.” Amazon must get packages to customers. Moving packages from warehouse to airport is easy. Flying them to Dubuque or Portland is straightforward, too, as is moving them from those airports to those shipping sites.

But now it gets complicated. We are to the last mile problem. Getting that box to your house (the last mile) is the complicated part of the equation.

Classroom implementation is the last mile in reading reform.

For instance, a major reform effort a decade ago created new state educational standards, an important lever. The new standards emphasized teaching kids to read texts of particular levels of complexity. More than forty states signed on, and publishing companies (another important lever) adjusted their reading programs accordingly.

But then the last mile. National surveys show that teachers persist in teaching with instructional level texts, instead of grade-level texts. So much for levers.

It isn’t just Jeff Bezos who should be losing sleep.

Your question about why it’s so hard to raise reading achievement points out the last mile problem in my opinion.

Imagine a veteran second-grade teacher, Ms. Jones. She’s always received good evaluations from her principals, the parents are happy to have their kids in her classroom, and whatever this or that test may say, she can see that her students make progress. They can read.

Now, the leveraging starts. We want that teacher to teach more phonics, or less. We want her to build knowledge instead of reading skills, or to work with harder books. Leveraging thrives on urgency, and its black-and-white rhetoric often sounds like, “If teachers don’t do what we say, kids won’t learn.”

But Ms. Jones has fifteen-years’ experience that tells her that the rhetoric is BS!

She doesn’t do whatever the leverage is touting, and yet she knows for a fact that her children are learning to read. Her own success is one brake on reform—why change if what you are doing is working?—but the overwrought rhetoric is a second. Why change if you can’t trust the people who are urging you to change?

Let’s face it. Our problem in reading isn’t that nothing works. It’s that everything does.

In the 1960s, researchers cooperated in a couple of dozen linked studies to determine what gives kids the biggest boost in reading achievement. They considered lots of possibilities: basal readers, phonics, programmed readers, linguistic readers, language experience approach, and so on.

The outcome?

All of those approaches worked and pretty equally. Oh, there were some differences—those that provided explicit decoding teaching did a bit better, as did those with a writing component. But, basically, everything worked.

Of course, these days a lot of those first-grade programs are obsolete. They’ve been replaced by reading workshops, guided reading, multiple cueing systems, decodable texts, research-based this, and child-centered that. And guess what? They all work, or at least to some extent they do.

Recently, School Achievement Partners released an analysis of Lucy Calkin’s Units of Study. I helped with that. We scrutinized the degree to which the program was in accord with the research on reading instruction, including how most effectively to serve English Learners. The response of many teachers who are using that program is that the research must be wrong because they know their students are learning. And they are. Just not as well as they could be or should be.

Learning to read in English is coming to terms with a writing system. That it is a system means that someone can figure it out. Instruction helps with this figuring out, but some kids are advantaged enough that they can do okay even with low instruction approaches.

The instructional research summarized by the National Reading Panel didn’t show that phonics instruction worked and that nothing else did, or that if you don’t get phonics, you’ll be illiterate. It showed that providing explicit phonics instruction in grades K–2 improved kids’ reading success. In other words, there were either fewer reading failures or marginally higher average achievement across the board.

The same is true for phonemic awareness, vocabulary, comprehension strategy, and fluency instruction. I promote teaching kids to read with grade-level texts instead of instructional-level ones, but not because the more demanding text regime ends with reading and the easy-text approach with failure. I’m clutching for the marginal advantage.

The U. S. is a highly literate nation. Almost all of us can read—no matter how we’ve been taught. But we’ve constructed a society around literacy. Reading is deeply implicated in our academic, economic, civic, and social lives. Achieving the levels of reading that we have in the past is insufficient. Ms. Jones has done well, but if today’s boys and girls only read as well as her students did a decade ago, they’re being disadvantaged.

That’s where Ms. Jones and the last mile become significant. As long as our rhetoric fails to correspond with her experience, we can lever all day long, but won’t deliver significantly higher reading achievement on scale because the last mile won’t be implemented.

The last mile rhetoric shouldn’t be a hair-on-fire message, but one that acknowledges both the current successes and the need to do better.

“Ms. Jones, we need your help. Studies show that kids can do better in reading if they receive a substantial amount of high-quality phonics instruction. Research also shows that hasn’t been happening in enough classrooms. We know you’ve been successful in teaching reading, but the goal line has moved. We need to get kids to higher levels than in the past, and that’s going to require some changes. Doing what we ask won’t change everything (and it’s not a criticism of your past efforts), but it will be better for your students and we all want that.”

Perhaps the strong rhetoric will move the levers, but remember we also have to persuade Ms. Jones in the last mile.

Achievement gaps between affluent and low-income students are caused by much more than what happens in the classroom. Poverty is associated with a litany of social consequences that make learning more difficult, such as unstable housing, poor healthcare, and greater exposure to violence and other traumas. Children burdened with these obstacles are less likely to attend school regularly, arrive on time, be prepared to learn, and more.

One approach to mitigating these harms and their effects on educational outcomes is the “community school.” The underlying idea is that, in addition to a school being a building in which children learn, it can also be a hub for social services that help disadvantaged families. Examples have been around in some form or another for over two decades, but they’ve become more popular in recent years, and more than 5,000 now operate across the country. Driving this in part are the strong claims of advocates as to the benefits of the community school. “Its integrated focus on academics, health and social services, youth and community development, and community engagement leads to improved student learning, stronger families, and healthier communities,” says the Coalition for Community Schools.

A recent RAND report examines how well these schools can deliver on that promise by studying their impact in New York City, where 267 of them currently operate—the largest collection in the nation. Launched in 2014 under Mayor Bill de Blasio through the New York City Community Schools Initiative, the effort enjoys strong governmental support and significant additional funding on top of standard allotments for academic instruction. In 2018, for example, when there were 239 such schools, they benefitted from $198.6 million in state, local, and federal dollars to provide community school services—including $118 million for academic enrichment efforts, $39 million to extended instruction time, and $18 million for health, mental health, and dental care.

RAND examined the impact of the program from its inception in 2014 through the 2017–18 school year. Using administrative and survey data provided by the New York City Department of Education, researchers compared the outcomes of students attending community schools to “a carefully constructed comparison group of schools that are similar to [them] in many ways, except for their designation.” The methodology used was “quasi-experimental,” meaning the schools in the treatment and control groups weren’t truly randomized. This was the best the researchers could do “because the initiative was launched at a large scale and focused on all New York City schools that failed to meet specific academic goals.”

RAND found that the initiative had a wide variety of positive effects. Among the more substantial are that New York’s community schools boosted attendance in elementary, middle, and high school; positively impacted on-time grade progression in elementary and middle school, as well as high school graduation rates; reduced chronic absenteeism, especially in schools with more substantial mental health programs; and in the final year studied, improved math achievement.

There were also a series of less substantial impacts that fall more on the side of social-emotional learning: a reduction in disciplinary incidents in elementary and middle school, but not in high school; teachers reporting a greater sense of shared responsibility for student outcomes, but only for two of the years studied, and not in high school; and students reporting a greater “sense of connectedness to adults and peers,” but only in elementary and middle grades, and only in one of the years studied.

Much of this is good, of course, as far as it goes, but it’s somewhat underwhelming, especially considering that these schools each receive upwards of $800,000 annually in additional funding. Despite students being in class more often and purportedly being more academically prepared to progress through grades on-time, there isn’t much change in their achievement. Yes, the math improvement in the 2017–18 school year is promising, but math was flat from 2015 to 2017, and reading was flat in all three school years. Moreover, the mental and emotional benefits seem limited.

To be sure, the RAND researchers are correct when they say that “The positive findings of the impact of the [New York City Community Schools Initiative] suggest that the strategy can be a promising approach to support student success in traditionally disadvantaged communities.” But they’re also right when they say that “New York City and its school system are unique—therefore, whether this initiative can be replicated elsewhere with similar findings is unknown.” Part of that uniqueness is the substantial amount of money the city has to allocate to such services.

Nevertheless, if other locales around the country want to rely on the community schools model, and they have enough funding to do so, this study suggests that disadvantaged students will benefit in measurable ways. They just ought to be sure there aren’t better ways to spend those dollars, and the time of teachers and other staff, that would benefit those children even more.

SOURCE: William R. Johnston, John Engberg, Isaac M. Opper, Lisa Sontag-Padilla, and Lea Xenakis, “Illustrating the Promise of Community Schools: An Assessment of the Impact of the New York City Community Schools Initiative,” RAND Corporation (2020).

Civics-education aficionados (and worriers) are generally acquainted with the 2018 issue brief from the Center on American Progress titled The State of Civics Education. It usefully summarized civics-education requirements and practices in the several states, highlighted a few worthy cases in point, and documented the parlous condition of civic learning in the U.S. at this point in our history.

Eighteen months later—December 14, 2019, to be exact—a different pair of CAP analysts uploaded a sort of update and extension of that brief, heavily focused on the “engagement” and “participation” elements of civic education and civic life, with an especially heavy emphasis on young people voting. (It’s hard to miss the connection to what CAP is hoping will happen in November 2020!)

The heavily-footnoted new paper, though short, makes several valuable contributions:

• It updates the data on state requirements in this realm. Thirty-nine states (and D.C.) require some sort of standalone civics course, and twenty states mandate some sort of civics exam for high school graduation.


• More originally, it stipulates five key elements of a “robust” civics curriculum and identifies how many of those are satisfied in every state. To my eye, the quintet, worthy as it is, is more like the beginning of a curricular skeleton than a robust body of knowledge, skills, and understandings, but that actually underscores how little is expected to be learned in this realm. (For what it’s worth, about half the states touch all five of CAP’s civics-ed bases.)


• We see that “community service” can earn graduation credit in twenty-three states, but that only Maryland and D.C. require it. (Which, in my view, may be just as well.)


• Only eleven states are seeing scores that average 3 or higher among kids who take the AP exam in American government—and that doesn’t correlate with how much civics education those states require (or how “robust” it is). Here, too, we find an example of the thinness of just about everything in the civics-ed space; as there’s nothing akin to a national assessment of civics other than NAEP, and as NAEP doesn’t provide state-level data on civics, the authors had to rely on AP exams, which aren’t representative of anything! (Of course, they opted not to reference the many states that are beginning to mandate the citizenship test.)


• The brief offers a new-to-me table showing voter participation (in 2018) in the eighteen- to twenty-four-year-old population—which ranged as high as 46.9 percent in Wisconsin and as low as 19.9 percent in Idaho.

Whereupon the authors launch into a series of earnest recommendations for boosting civic activism, civic participation—and, yes, voting—among young people, some of which I find credible and worthy, others not so much.

See for yourself. The wave of interest around the country in better civics education is commendable. But the line between “knowing” and “doing” is getting blurrier, and some of the latter is growing harder to distinguish from the political agendas of those encouraging it.

SOURCE: Ashley Jeffrey and Scott Sargrad, “Strengthening Democracy With a Modern Civics Education,” Center for American Progress (December 2019).

On this week’s podcast, Andrew Campanella, president of National School Choice Week, joins Mike Petrilli and David Griffith to discuss whether affluent suburban kids deserve vouchers too. On the Research Minute, Amber Northern examines the extent to which older siblings affect college-going behavior. 

Amber's Research Minute

Joshua Goodman, Michael Hurwitz, Christine Mulhern, and Jonathan Smith, “O Brother, Where Start Thou? Sibling Spillovers in College Enrollment,” Retrieved from Annenburg Institute at Brown University (December 2019).