1. The Dispatch took a gander at which educational institutions—including K-12 schools—in central Ohio were approved for federal PPP loans. (Columbus Dispatch, 7/8/20)
    
2. A little closer to the ground, here’s a story about a nonprofit organization providing laptops—and internet access and software and educational coaches—for low-income families in a number of states including Ohio. The main point is that students who are not yet in Kindergarten need that experience with technology to help them be ready—and that’s the program’s raison d'être—but it has had some larger effects too. We learn about the Massey family here in Columbus, who were able to boost the amount of device time available for the whole family with the new laptop donation for the youngest child, enabling everyone to participate in summer learning. It also gave the family the capacity and confidence to opt for fully online learning for next school year for all of the children. Which seems like just what Mom wanted all along. (Spectrum News 1, 7/7/20)
   

Did you know you can have every edition of Gadfly Bites sent directly to your Inbox? Subscribe by clicking here. 

As the start of the school year rushes toward us, millions of teachers across America are girding themselves for their new role as “essential workers” during a persistent pandemic, looking ahead with trepidation to teaching while wearing masks, behind Plexiglas, and to students who will have been out of the classroom for six months or more.

As dreadful as all of that is, one group of teachers has it particularly rough: U.S. history instructors. They must also perform their duties during a full-scale culture war over how to tell the American story, especially on the central and complex issue of race.

True, questions about how best to teach American history have been fraught—and fought—for decades, certainly since the rise of the revisionists and the publication four decades back of Howard Zinn’s A People’s History of the United States. Particularly sensitive is how to handle America’s original sins of slavery and white supremacy. But with the New York Times’s controversial 1619 Project winning the Pulitzer Prize this year, youthful mobs toppling statues, and a series of counterpunches from President Trump, the debate has jumped from teachers’ lounges to the front pages.

This cannot be good for high school history courses, or for kids who are both required to study U.S. history and ought to understand it.

Imagine you live in deep-red Alabama. Can you picture teaching an American history that veers from the deep-red doctrine? Imagine you live in deep-blue Boston. Can you picture teaching an American history that veers from the deep-blue doctrine? Is it even possible to teach a purple version of American history? Does such a thing even exist?

—

How things have changed from just five years ago, when Hamilton’s smashing Broadway success pointed to an inclusive, inspiring way to grapple with the contradictions at the heart of America’s founding. Jonah Goldberg summarized it beautifully in his chapter in our recently-published volume, How to Educate an American:

Only fools and bigots could belittle [slavery’s] evil. Yet slavery’s resonance in America comes not from its evil but from the founders’ hypocrisy. Since the Agricultural Revolution, nearly every civilization practiced slavery, but none also claimed to believe that “all men are created equal” and that they are “endowed by their Creator with certain unalienable Rights….” The staggering hypocrisy of slavery is regrettable in one sense, but glorious in another. Hypocrisy is only possible when you have ideals.

As Bret Stephens wrote this week, that’s the story of U.S. history as “gradually and imperfectly unfolding liberty.” But the revisionists—most notably, and recently, Nikole Hannah Jones of the 1619 Project—instead promote a story “of unbending oppression.” Or, as Goldberg puts it, “The Zinnian approach takes America’s sin of slavery and makes it an eternal blemish that never shrinks in the rearview mirror no matter how much progress we make.”

Making matters worse is when the revisionists stretch the truth to fit their narrative. Imagine how much better received the 1619 Project would have been among conservatives had Jones not propagated the lie—that really is the best word—that “one of the primary reasons the colonists decided to declare their independence from Britain was because they wanted to protect the institution of slavery.” (This was later amended by the Times to “some of the colonists.”) Jones could have simply laid out the gripping and troubling history of 1619 and argued that it should be studied by schoolchildren rather than swept under the rug. But instead she targeted the miraculous accomplishments of 1776 as well, and did it with dishonest attacks.

Surely Jones’s take on history—and Zinn’s—is what Trump was referring to when he said at Mount Rushmore that

…our children are taught in school to hate their own country, and to believe that the men and women who built it were not heroes, but that they were villains. The radical view of American history is a web of lies—all perspective is removed, every virtue is obscured, every motive is twisted, every fact is distorted, and every flaw is magnified until the history is purged and the record is disfigured beyond all recognition.

And was it really so controversial when he argued that:

We must demand that our children are taught once again to see America as did Reverend Martin Luther King, when he said that the Founders had signed “a promissory note” to every future generation. Dr. King saw that the mission of justice required us to fully embrace our founding ideals. Those ideals are so important to us—the founding ideals. He called on his fellow citizens not to rip down their heritage, but to live up to their heritage.

Above all, our children, from every community, must be taught that to be American is to inherit the spirit of the most adventurous and confident people ever to walk the face of the Earth.

Americans are the people who pursued our Manifest Destiny across the ocean, into the uncharted wilderness, over the tallest mountains, and then into the skies and even into the stars.

We are the country of Andrew Jackson, Ulysses S. Grant, and Frederick Douglass. We are the land of Wild Bill Hickock and Buffalo Bill Cody. We are the nation that gave rise to the Wright Brothers, the Tuskegee Airmen, Harriet Tubman, Clara Barton, Jesse Owens, General George Patton, the great Louie Armstrong, Alan Shepard, Elvis Presley, and Mohammad Ali. And only America could have produced them all.

In fact, it sounds rather similar to what Joe Biden said in the aftermath of the George Floyd killing, in a speech I lauded as an example of how to teach a patriotic, if imperfect, version of American history:

We build the future. It may in fact be the most American thing to do.

We hunger for liberty the way Harriet Tubman and Frederick Douglass did.

We thirst for the vote the way Susan B. Anthony and Ella Baker and John Lewis did. We strive to explore the stars, to cure disease, to make this imperfect Union as perfect as we can.

We may come up short—but at our best we try.

In other words, both Trump and Biden seem to be saying, let’s tell our story like Hamilton does. 

Yet Hamilton, less than five years old, is now considered “not woke enough” in some quarters. Doesn’t that strike folks as crazy? As Megan McArdle wrote several weeks ago in reference to the ongoing statue-cide, it’s critical to be able to draw lines, to make distinctions, to say when enough is enough. Otherwise the mob rules, and we sacrifice our precious inheritance—of history, of the Enlightenment, and of the truth.

—

So back to our beleaguered history teachers. It would be hard to fault them for trying to sidestep any and all controversies this fall, especially in light of recent injustices and racial tensions, or to take the edges off them with bland, antiseptic readings, such as those common in textbooks. But that well-traveled road leads to bored, disengaged students, and contributes to young Americans’ woeful knowledge of our nation’s history.

A better approach is one modeled by the new A.P. U.S. History curriculum, which, after some initial missteps, manages to tell the American story in all its fullness and glory without slipping into politicization or veering from the truth. Other courses could as well.

Let’s be honest though: Such an approach won’t make everyone happy; it might not make anyone happy. So we should remember to applaud our history teachers who show the courage to do it anyway, who give our young people the complete picture of America, not just one that fits the preferred narratives of the left or the right. Because the courage to speak the truth is part of the American story, too.

Editor’s Note: The Thomas B. Fordham Institute occasionally publishes guest commentaries on its blogs. The views expressed by guest authors do not necessarily reflect those of Fordham.

Author’s Note: The School Performance Institute’s Learning to Improve blog series typically discusses issues related to building the capacity to do improvement science in schools while working on an important problem of practice. Now, however, we’re sharing how the United Schools Network, a nonprofit charter management organization based in Columbus, Ohio, is responding to the Covid-19 pandemic and planning educational services for our students during school closures.

As schools consider how to restart in the fall, some are galvanizing their community to consider this an opportunity for reinvention, a way to rethink how we educate our young, while others simply want to get students back in the building and return as quickly and safely as possible to the normal that existed prior to the pandemic. Regardless of a school’s position on the spectrum from restart to reinvention, a more sophisticated manner in which to analyze student engagement and track efforts to improve engagement will be necessary.

In prior blog posts, my colleague John A. Dues explored the idea that successful remote learning requires a whole new system, especially because many teachers were accustomed to in-person teaching and had no prior experience with teaching students from afar through a computer. I followed up with some thoughts on defining student engagement by creating operational definitions for key remote learning concepts (like engagement) and then using those definitions to inform data gathering. Finally, John discussed the use of Process Behavior Charts (PBCs) to better understand and analyze student engagement data, and how to use a basic knowledge of variation to avoid common pitfalls of under- and overreaction to these data. Let’s pick up the conversation there.

Inspired by the legendary management consultant W. Edwards Deming and the esteemed statistician Dr. Donald J. Wheeler, Mark Graban, in his book Measures of Success, proposes three key questions that organizations must consider when using data to drive improvement:

1. Are we achieving our target or goal?

a. Are we doing so occasionally?

b. Are we doing so consistently?

2. Are we improving?

a. Can we predict future performance?

3. How do we know we are improving?

a. When do we react?

b. When do we step back and improve?

c. How do we prove we’ve improved?

The engagement data we collected from Columbus Collegiate Academy-Main Street during the first two weeks of remote learning is displayed in the table below. Our operational definition for engagement is: “A USN middle school student demonstrates engagement in a remote lesson by completing the accompanying practice set in its entirety.”

​

Let’s apply Graban’s key questions to the data table.

1. Are we achieving our target or goal? We set our target as 70 percent engagement, considering we were beginning a brand new system of instruction and faced significant barriers to successful implementation (internet connection, access to devices, etc.). It appears that we are hitting 70 percent engagement in some classes on some days.

a. Are we doing so occasionally? Yes.

b. Are we doing so consistently? It is hard to tell from the data. While the color-coding draws attention to specific numbers, it is hard to tease out patterns in the data presented in this way. In addition, the color-coding, which corresponds to our grading scale, does not easily communicate achievement of our target (70 percent is still in the “yellow”).

2. Are we improving? It is hard to tell from the data. The simple list of numbers makes it difficult to identify patterns.

a. Can we predict future performance? Not from the data presented in this manner.

3. How do we know we are improving? We do not know if we are improving.

a. When do we react? We cannot know by presenting the data in this way.

b. When do we step back and improve? We cannot tell by presenting the data in this way.

c. How do we prove we’ve improved? We cannot prove we’ve improved by presenting the data in this way.

The table above is a start, and certainly better than not collecting and examining the data at all. The colors allow you to quickly obtain a very broad picture of remote learning engagement. More green is good, more orange/red is bad. However, the data, as presented in this way, cannot speak to improvement. A Process Behavior Chart (PBC) is a better option.

The PBC below takes data in a table and, by “plotting the dots,” allows you to see change over time in a more efficient manner—like a run chart on steroids. While the details on creating PBCs are beyond the scope of this blog post, I’ll quickly review what we see below.

The red lines are the upper and lower limits. You should expect engagement percentages to fall within these limits based on the data set. The green line is the mean (or average), and the blue dots are the engagement percentages for the given day. This PBC, for eighth grade math specifically, is telling us that we can expect remote learning engagement to fall between 47 and 85 percent on any given day, with a mean engagement rate of 66 percent.

​

Graban proposes ten key points when utilizing PBCs to make data-driven decisions. Most importantly, he states that we manage the process that leads to the data we see above, not the data points themselves. We can spend hours blaming people for metrics we do not like, but that time is largely wasted. In the context above, the “voice of the process” is telling us that, if we don’t do a single thing to change the system, we can expect similar results moving forward—daily engagement between 47 and 85 percent. If we do not like the limits or the mean, we need to go to work on the process. Furthermore, we must understand the basic reality that variation exists in every data set; if we react to every “up and down” (called “noise”), we will likely burn out and miss actual signals in the data that we can use to improve. Why did engagement increase 8 percent from March 19 to March 20? Who knows? This is likely unknowable, and not worth further exploration. So when do we react?

In order to better analyze the PBC above, let’s apply Graban’s three rules:

• Rule 1 Signal: any data point outside of the limits

• Rule 2 Signal: 8 consecutive points on the same side of the mean

• Rule 3 Signal: 3 out of 4 consecutive data points closer to the same limit than the mean

​

As demonstrated above, we see one example of each rule in our 8th grade math remote learning PBC. We can use this learning to modify the PBC to make it easier to determine next steps on the path to improvement.

Below, after recognizing the “run of 8” that indicated engagement was increasing for a sustained amount of time, we adjusted the limits and mean for that specific chunk of time. Interestingly, the start of this improved period aligned with our return from spring break. We then adjusted the limits and mean after a sustained period of improvement when we detected a downward turn. Thus, a PBC that began as a singular, predictable system with wide limits has been redefined to a system with three distinct periods of varying engagement. An initial stage of highly-variable engagement (which makes sense, considering remote learning was brand new to everybody) gave way to a period of increased engagement (possibly due to students and teachers becoming accustomed to the new reality), and finally a period of decline as we approached the end of the school year.

​

One can imagine how schools could utilize PBCs to make more informed decisions on interventions, either proactively or retroactively. At USN, like many schools, we had to adjust to remote learning extremely quickly, and therefore did not have the time or bandwidth to track interventions with as much sophistication as we would’ve liked. However, let’s simulate how we could utilize PBCs moving forward, especially in light of the fact that remote learning, in some form, may continue into next year.

​

Eighth grade math experienced two significant shifts this spring in the remote learning context. Looking backward, with the goal of improving remote learning next year, we could ask ourselves: What was different between April 20 and May 7? Was the teacher texting students daily to remind them to complete assignments? Did we improve the way in which instructional materials were organized in Google Classroom to make it easier for students to find and complete them? Along the same lines, we see another unique system between May 8 and the end of the school year. What about our system changed that led to decreased engagement? At times, these questions may be difficult to answer, but when we detect a signal (remember the three rules above), it’s time to grab our shovels and dig.

This approach works proactively, as well. For example, we can pretend that today’s date is April 8, and we’re planning an intervention such as eighth grade teachers will text every student every day to remind them to complete their remote assignments. We collect daily data, and post-intervention, we note increased engagement. We now have data to demonstrate that this intervention shows promise! Conversely, pretend that today’s date is May 8, and we’re planning another intervention to try to increase engagement even more. We might ask eighth grade teachers to plan twice-weekly Google Hangouts to provide support to students who sign on. We collect the data for a few weeks and, unfortunately, the intervention does not lead to improvement. Moreover, average engagement decreased. While that’s a bummer, we still learned something. Above all, it’s vital to note that the focus is on the process—the changes we introduce to the system to improve it—not the people.

As a final step, let’s revisit our three key questions once more using the PBC instead of the table of numbers.

1. Are we achieving our target or goal? As a reminder, we set our target as 70 percent engagement. On average, we are not achieving our goal in the subject of math, with the exception of a two-week period following spring break.

a. Are we doing so occasionally? Yes.

b. Are we doing so consistently? No.

2. Are we improving? We are not improving. We experienced a period of improvement following spring break, but that was followed by a period of decline toward the end of the school year.

a. Can we predict future performance? We currently have a stable system with the exception of one signal on the last day of school. We can predict that engagement will vary between 45 percent and 73 percent, with a mean of 59 percent, unless something substantial shifts in our system.

3. How do we know we are improving? We can apply Rules 1, 2, and 3 to the PBC to separate signals from noise and determine if, and when, improvement occurs.

a. When do we react? We can react after we see a “run of 8” in the data, signaling an improvement in the system. We also reacted upon examining a Rule 2 later on that signaled a decline in engagement.

b. When do we step back and improve? After a period of improvement in mid-April to early May, we ended the year with a period of decline. We can use this learning to think about designing a better remote learning system next school year.

c. How do we prove we’ve improved? In this particular data set, we can prove we improved for a period in mid-April to early May as indicated by an increase in the mean engagement.

At SPI, we have much to learn about using PBCs in the field of education. However, we believe there is tremendous power in examining and acting on data in this way. We spend less time overreacting to (or underreacting to) common day-to-day variation, and more time improving the system and verifying the outcomes of interventions. Whether remote or in person, this type of data analysis will certainly help guide us in making smart decisions during an uncertain time.

Seventeen long years ago, I urged the creation of “religious charter schools,” either encouraging their start from scratch or—more realistically—allowing extant Catholic and other faith-based schools to convert to charter status without having to forego the religious elements that distinguish them and that many parents crave for their children.

With hundreds more Catholic schools having perished in the interim, that idea seems more important than ever—and with the Supreme Court’s Espinoza decision now in hand, the biggest single obstacle to realizing it may have fallen.

As Josh Dunn wrote recently in Education Next:

The Supreme Court’s decision siding with religious-school parents in Espinoza v. Montana Department of Revenue will have implications far beyond Montana. Though the court did not explicitly say so, its decision effectively struck down Blaine amendments, provisions in thirty-eight state constitutions forbidding aid to religious institutions. If the narrow 5-4 majority of the Espinoza decision remains intact, the case will be both one of the most significant religious liberty and education decisions in decades.

If states’ Blaine amendments are now headed to the ash heap of history, imaginative leaders of Catholic schools should be able to team up with smart Constitutional lawyers and like-minded legislators to craft ways to do this, most likely at the outset in “red” states.

Justice Breyer may have anticipated such a possibility himself, back during January’s oral arguments, though he framed it with blue jurisdictions in mind. As reported by Education Week:

“Say in San Francisco or Boston or take any city or state, and they give many, many, many millions of dollars to the public school system, and a lot of them give a lot of money to charter schools,” Breyer said. “Now, they don’t give money to Catholic schools. All right? Now, if we decide you’re right, does that all change?”

Yes, it could now change, say I, and I believe Professor Dunn and other analysts would agree, though the odds of it actually happening in blue states such as California and Massachusetts are slim. Indeed, the likelier prospect in those places is a major rift within the choice and charter movements, as more than a few charter partisans want no part of religious schooling.

For those whose memories may have been disturbed by other recent events, “Blaine amendments“ refer to provisions in most state constitutions—nearly all adopted in a frenzy of anti-Catholic and anti-immigrant sentiment in the late nineteenth century and named for a prominent GOP congressman (and 1884 Presidential candidate)—that bar state aid to religious organizations of various kinds. For more than a hundred years, and despite sundry exceptions in other realms, they’ve pretty much blocked direct state and local subsidies to parochial schools of all kinds across much of the United States. Never mind that such support is taken for granted in almost every other modern land.

The First Amendment to the U.S. Constitution, however, throws a curve ball at such provisions because, in addition to barring the “establishment of religion,” it vouchsafes to Americans the free exercise of religion. That’s what tipped the balance for five out of nine justices in Espinoza, ruling that “A State need not subsidize private education. But once a State decides to do so, it cannot disqualify some private schools solely because they are religious.”

Similarly, seems to me, a state need not create charter schools—and five have so far opted not to—but if it does allow (and pay) for their creation, can it constitutionally disqualify some because they are or hope to be religious?

I’m no attorney. But this door looks to be worth opening.

Here’s my original proposition, whose time, I hope, may now finally have come:

— —

The tragedy of urban education is the dearth of effective schools for poor kids. That acute shortage belies the right nominally conferred by the federal No Child Left Behind Act, namely that parents can move their children from failing public schools to better ones. Many communities have nowhere near enough capacity in well-functioning schools to provide an education haven for those thousands of youngsters. (In cities like New York and Chicago, we’re talking hundreds of thousands.)

Federal law also says such kids may go to charter schools, but there aren’t enough of them, either, at least not the highly effective kind.

How to get more? Take advantage of the charter option and become more creative and open-minded. Many cities with weak public schools have strong churches and faith-based organizations. And one thing that many parents crave for their children is a school that not only teaches the three R’s, not only keeps Tony and Tanika safe, not only gives them a teacher who knows their names and cares if they’re learning—but that also supplies them with values, morals, a code of behavior, and a sturdy faith in God.

Yet the No Child Left Behind legislation doesn’t include the right to go to private schools, where such things are routine. Paul G. Vallas, the chief executive officer of the Philadelphia school system, is seeking a way around that restriction, hoping to send disadvantaged youngsters from troubled public schools into archdiocesan classrooms that have space and are willing. But, like vouchers, this is an uphill political battle. And even with voucher aid, many children who would benefit from the curricular and moral offerings of private schools cannot afford to matriculate. But faith-based organizations seeking to operate zero-tuition charter schools have, until now, been compelled to exclude all forms of religiosity—thus quashing one of their major incentives to serve children and barring one of the things they do best.

Solution: Let religious schools become part of the charter system so long as they’re willing to abide by the results-based accountability arrangements that other charter schools must accept, namely state academic standards and tests. And allow churches to found new charter schools without shedding their sectarianism.

In most countries, this wouldn’t qualify as an innovation, for they assume that government has an obligation to support church-affiliated as well as secular schools. In the United States, however, a daft reading of the First Amendment’s “establishment” clause was long held to bar public aid to religious schools.

The U.S. Supreme Court’s 2002 Zelman decision changed this. It said there’s no federal prohibition on state dollars going to such schools so long as this results from free and open choices by parents. It thus legalized voucher programs in several states and others struggling to be born, including the new one in Colorado and the District of Columbia plan now pending in Congress.

But vouchers aren’t the only education innovation that Zelman makes possible. Charter schools, too, get public dollars only when parents freely choose them. No child is compelled to attend a charter school. If parents don’t select it, it has neither pupils nor revenue.

Yes, charter schools must be “sponsored” by state-approved agencies, and some will see excessive “entanglement.” But private schools also need state licenses and, under the pending District of Columbia voucher program, must accept other constraints devised by Congress. Doing so will not, however, erase their religious character.

True, other differences remain between private and charter schools. Wholly private schools can restrict attendance to members of their faith and expel youngsters who refuse to behave. They can operate separate programs for girls and boys and need not comply with Uncle Sam’s myriad rules for educating children with disabilities.

Because they value such freedoms, many private schools will shun greater involvement with government. So be it. But some would welcome the opportunity to serve more children. In most places, per-pupil funding for charters, meager as it is, exceeds current tuition levels—and is more than vouchers would bring. In any case, the entanglements that accompany charter school status are not much worse—as is becoming clear in Florida, where new rules are raining down upon private schools that take part in that state’s several voucher programs. It’s a calculation each school can make for itself.

—

A few private schools have converted to charter status, but they did so by severing all religious ties. A few others have created sister schools that operate as charters. I visited an interesting pair of schools in Houston, one private (and religious), the other charter (and secular), sharing facilities but functioning as separate organizations.

Creating a secular sister school is one viable model for a parochial school or church that wants to serve more kids. It may be the only option in states with “Blaine amendments” that prohibit public dollars from flowing into religious institutions no matter what the Supreme Court says about the U.S. Constitution. But in the dozen or so states without such impediments, why not try religious charter schools?

Watchdog groups will rush back to court at the first sign of a new breach in their cherished “wall of separation,” and in time this education innovation would also wind up in the Supreme Court. But that’s no reason to forgo it. Cash-strapped states may fear the budgetary impact of private school pupils suddenly qualifying for public subsidies. Yet that cost can also be contained. Since many state charter laws bar private school conversions, most religious charters would be new schools, serving kids not already in the private school orbit—and adding to the supply of seats in decent schools for youngsters who need them.

Overriding all objections is America’s woeful lack of such seats. Every possible asset should be brought to bear on the creation of more. Religious charter schools deserve consideration.

The original 2003 article first appeared in Education Week.

Last week, Bellwether Education Partners released three briefs examining the past, present, and future of modern school accountability systems.

Our goal with these briefs was to reflect on the status of the standards-based reform movement. As originally conceived, equity was at its core, reflected in a commitment to high, universal expectations for all students and consequences for failing to close achievement gaps between students with different racial, income, and learning characteristics. How well has the underlying theory of standards-based accountability held up? Given what we’ve learned from more than two decades of successes and failures, how should these systems continue to evolve?

We start by examining what’s worked and what hasn’t. Based on the most rigorous research, accountability policies produced measurable improvements, particularly for traditionally underserved student groups. But those gains have not been substantial enough to close achievement gaps as their architects hoped. In particular:

• They’ve led to gains in math test scores, particularly for historically disadvantaged students in younger grades, but have not led to discernible improvements in English language arts.
• They’ve contributed to a real rise in high school graduation rates, yet achievement scores in the upper grades have not risen commensurately.
• They’ve produced a trove of transparent data for research and evaluation, but that information is often not intelligible to parents or actionable for teachers.

Rather than accepting these as universal truths about “accountability” systems writ large, we found it useful to break down the theory of action into its component parts. Here, in brief, are our five diagnoses.

1. Learning standards

Setting clear, rigorous learning expectations for all students is essential to the theory of action behind school accountability. While states have made progress on this front in the past decade, more work remains to ensure that academic standards lead to better and more equitable student outcomes. Moreover, while clearer and higher standards can create the conditions for student success, high-quality standards have not been powerful enough on their own to ensure equitable access to a quality education.

2. Annual summative assessments

States have a solid track record of producing reliable and comparable data on every student’s progress toward state standards. Yet that doesn’t mean state testing programs are realizing their potential. Many states fail to deliver assessment results in a timely manner, limiting the ability of parents and advocates to use assessment data to inform supports or policy changes to address student needs.

3. Transparent, disaggregated data

Gains in transparency regarding the performance of schools and students, particularly disparities in outcomes for traditionally underserved students, are among the biggest wins of the modern accountability movement. Parents and advocates gained a powerful tool once states began providing performance data disaggregated by student subgroups. Data from state accountability systems have also helped fuel a new wave of worthy research on education policies and practices.

But while all these data have been good for advocates and analysts alike, simply making data available has had no positive effect on how teachers or schools behave. Research has, however, shown that transparency alone is not enough to change student outcomes.

4. School ratings

After testing, the most prominent mechanisms of standards-based accountability systems are school ratings. These serve two formal functions: to assign each public school a clear label that reflects its performance, and to identify schools with low overall or subgroup performance as a marker of the need for additional support and interventions.

School ratings have been shown to drive positive performance at least in the short run, but they may not be enough to boost enduring gains in student outcomes. Moreover, the value of school ratings is tied to the measures used to create them, which can encourage schools to focus on the wrong things—such as prioritizing students on the bubble of meeting a proficiency target rather than ensuring that every pupil makes sufficient growth each year.

5. Interventions in low-performing schools

Once policymakers identify low-performing students and schools, they can target resources and interventions accordingly. But the threat of interventions had more impact that the interventions themselves. As a result, when enacting ESSA in 2015, Congress scaled back the specificity and scope of federally imposed interventions. Today, Washington still requires interventions in low-performing schools, but states and districts are free to determine what those interventions will be and what supports, if any, to provide.

This may be the right balance, but dramatic improvements in student outcomes require dramatic changes to all or some parts of a school’s curriculum, instruction, and staffing. We don’t hold out much hope that all states are going to follow through with strong interventions. Some will, but many more jurisdictions will simply default to procedural “needs assessments” rather than measures leading to real changes.

—

States are already proposing to cancel their assessments for the spring of 2021, but we think those declarations are premature and misguided. While there is little consensus on what schooling will look like in 2020–21, what is certain is the massive risks to student well-being and academic progress that will come as a consequence of the pandemic and the related economic fallout. As detailed more extensively in our third brief, the staggering learning losses projected as a consequence of Covid-19-induced school closures threaten to widen achievement gaps even further.

The massive budgetary, educational, and health challenges schools face make it even more important to rediscover the common ground that accountability systems once held. We can’t merely go back to the way things were, and we shouldn’t. These systems weren’t perfect, and a forthcoming webinar with Jeb Bush, John B. King, Jr., and Carissa Moffat Miller will explore how they must evolve to address key challenges and shortcomings.

But the theory of action behind standards-based reform—setting standards, measuring progress, and applying mechanisms of accountability for outcomes to promote equity for every student—is sound and remains as relevant as ever. To ensure that students aren’t left behind as schools continue educating during a pandemic, policymakers must ensure that accountability systems are adapted and sustained rather than sidelined during this critical moment for kids.

Early reporting of suspected maltreatment is one key to mitigating the damage that abuse can inflict on a child. Yet with millions of school age children having their classroom time shortchanged due to the coronavirus, a primary source of detection for maltreatment has been cut off—teachers. Service providers are justifiably concerned with drops in the number of abuse referrals, as are medical professionals with the increased severity of abuse cases they have witnessed in recent months. A timely report from NBER reiterates just how vital schools are in detecting and breaking these horrific abuse cycles.

Statistics show that around 13 percent of all children in the U.S. have a confirmed case of maltreatment by age eighteen. According to federal law, maltreatment is “any recent act or failure to act on the part of a parent or caretaker which results in death, serious physical or emotional harm, sexual abuse or exploitation; or an act or failure to act, which presents an imminent risk of serious harm.”

The NBER analysts use two separate regression discontinuity designs. The first uses data from thirty-five states collected between 2003 and 2015 about child maltreatment as reported to the National Child Abuse and Neglect Data System (NCANDS) Child File. Specifically, these are referrals made to Child Protective Services (CPS) that are assessed to see if they merit further examination and, if so, whether they can be substantiated. Over that period, there were roughly 48 million reports of child maltreatment. In over 90 percent of cases, the child’s parent was the perpetrator. The most common form of abuse is neglect, followed by physical abuse.

The report utilizes statewide policies across states and over time regarding the earliest age at which a child is allowed to enter school. Researchers are testing whether the investigated reports of child abuse for five-year-olds to CPS, in particular, are greater for children born just before the statewide school entry eligibility cut-off date (they enroll in kindergarten in the year they turn five) relative to those born just after the entry eligibility cut off (they enroll in the next year). In other words, the study design leverages the fact that a child’s exposure to educators varies during the child’s fifth year of life based on being born on adjacent days. Still, schools and districts do not always follow these age-based admissions policies to the letter, so the results are essentially “intent-to-treat” estimates.

The headline is that they find an increase between 5.4 and 9.2 percent in reported instances of child maltreatment. Assuming that this is all new reporting (a different child on each report), that means 0.3 percent of five-year-olds are being newly reported to CPS agencies because of their kindergarten eligibility. They look at the types of reporters and show that school-based contact increases the number of first reports for a child, as well as the number of additional reports that confirm prior cases of suspected abuse.

The second discontinuity analysis uses variation in school calendars (annual start and end dates) set by districts to identify effects of exposure to school settings on child abuse reporting. This is similar to the first analysis, as the researchers are testing whether the number of reports is greater when children are attending school compared to dates when they are not. They use data from the twenty-five largest districts in the country, gleaning start and end dates from 2005 through 2016. They include all children from ages six to seventeen. In a nutshell, they find that reports to CPS go up between 48 and 64 percent at the start of the school year and also peak one month before the school year ends—almost as if educators are responding to the fact that students will soon be out of their sight for three months—but there is a decrease when school actually ends. While educators are responsible for between 33 and 41 percent of CPS reports in this analysis, about half of them originate from social service, mental health professionals, or others. The researchers suggest this is due to the wider age range of the sample, with older children having more contact with other adults on high school campuses.

The report reiterates that teachers play a key role in reporting child abuse, and rightly calls for their proper training to identify abuse, given the very uneven reporting across the country and in school districts. But the more urgent question poses an uncomfortable trade-off between the health of all children in schools and, in particular, those at risk of abuse: What if many schools don’t reopen in the fall due to ongoing coronavirus concerns?

SOURCE: Maria D. Fitzpatrick, Cassandra Benson, and Samuel R. Bondurant, “Beyond Reading, Writing, and Arithmetic: The Role of Teachers and Schools in Reporting Child Maltreatment,” NBER Working Paper #27033 (April 2020).

Enacted in 2012, Ohio’s (well-named) Third Grade Reading Guarantee aims to ensure that children can read proficiently by the end of third grade. The legislation calls for interventions when students struggle to meet early literacy benchmarks, and it includes a requirement that schools retain third graders who do not meet targets on state English language arts exams or an alternative assessment. At the bill-signing ceremony, former Governor John Kasich captured the logic behind the law saying, “If you can't read you might as well forget it. Kids who make their way through social promotion beyond third grade, when they get up to eighth, ninth, tenth grade…they get lapped, the material gets too difficult.”

Despite the vital importance of early literacy, the reading guarantee has faced significant pushback from the start. Critics have attacked the policy as an “unfunded mandate,” and they’ve assailed the retention provisions as a form of punishment. The latter argument was recently echoed by a state board of education member who justified her opposition to raising the promotional score on third grade state exams by saying, “I don't think that anyone should be punished more than they already are being punished.”

What the critics seem to forget is the data showing that third grade tests scores are highly predictive of later outcomes. Kasich was right. Children who have difficulty reading early in life are going to face immense challenges in middle and high school—and likely into adulthood. That’s the real punishment.

The most influential report showing the link between early literacy and later outcomes is a 2011 Annie E. Casey Foundation study by Donald Hernandez. His work is based on the National Longitudinal Survey of Youth which tracked nearly 4,000 children born between 1979 and 1989 until age nineteen. That survey collected data on third grade reading scores on the Peabody Individual Achievement Test and high school graduation.

The results are startling. Non-proficient readers in third grade are four times more likely to fail to graduate than their peers who read proficiently. The odds of dropping out rise among non-proficient Black and Hispanic students who are a staggering eight times more likely to not graduate from high school.

Hernandez’s study tracked students who were in elementary school during the 1980s and ‘90s. Perhaps the link between early literacy and later outcomes has dissipated in more recent years. Or maybe state exams aren’t very good predictors of later outcomes.

In comes a new study by the University of Washington’s Dan Goldhaber and Malcolm Wolff and EdNavigator’s Timothy Daly, which shows that third graders’ state test scores still strongly predict high school outcomes. Their analysis relies on student data from Massachusetts, North Carolina, and Washington State that allow them to track academic progress from grades three through twelve between 1998 and 2013.

They find that lower third grade math and reading scores are associated with significantly lower high school math scores (results in English aren’t reported), less participation in advanced coursework, and lower rates of high school graduation. In terms of exam outcomes, the authors put the results this way: “All else equal, a student at the lowest percentile of the third grade math test distribution rather than the highest percentile is expected to be 48–54 (depending on state) percentile points lower in the high school math test distribution.” Depressingly, the relationship between third grade results and high school outcomes is nearly as strong as eighth grade scores, suggesting that very little “academic mobility” occurs between third and eighth grade.

To illustrate the pattern visually, the figure below, taken from their report, displays the link between students’ third grade test scores on the x-axis and their likelihood of graduation on the y-axis. As is evident from the data points at the far left of the chart, students scoring in the lowest deciles—the first and second, especially—on third grade exams have much lower probabilities of high school graduation. For instance, for low-income students, defined by students’ eligibility for free and reduced-priced lunch (FRL), a widely used proxy for student poverty, the odds of graduation among those scoring in the lowest decile on third grade reading exams are about 0.70. That’s below low-income students who scored in the fifth decile, whose probability is just over 0.80, and far lower than high-achieving low-income third graders.

Figure 1: Probability of Graduation by third grade test score decile and FRL

​

Source: Dan Goldhaber, Malcolm Wolff, and Timothy Daly, Assessing the Accuracy of Elementary School Test Scores as Predictors of Student and School Outcomes, CALDER (May 2020).

Goldhaber and colleagues sum up their analyses with the powerful conclusion (emphasis added):

[E]arly student struggles on state tests are a credible warning signal for schools and systems that make the case for additional academic support in the near term, as opposed to assuming that additional years of instruction are likely to change a student’s trajectory. Educators and families should take third grade test results seriously and respond accordingly; while they may not be determinative, they provide a strong indication of the path a student is on.

State and local leaders in Ohio should heed the advice of these researchers and take third graders’ test scores seriously. This will be all the more important in the coming year, as students return to school having lost academic ground after being out of the classroom for an extended period. Although state retention provisions were waived for last year’s third graders due to the pandemic, schools must still step in and ensure that struggling readers receive the time and help needed to read proficiently. As the data indicate, these students can’t afford to get lost in the shuffle—a real possibility in the remote and blended-learning environments that schools are likely to use next year. Some critics will continue to wrongly decry such interventions as punitive. But let’s all remember that these steps are no punishment, but rather an investment in a child’s future.   

Editor’s note: This post was first sent as an email in The Bulwark’s newsletter “The Triad: Three things to read, from JVL.”

1. School?

Right on schedule, the national school panic is here.

Most of the country is supposed to go back to school in September. Parts of the country go earlier. Nobody has any idea what they’re doing or what’s actually going to happen.

Let’s lay out some guide rails for how to think about schooling in the pandemic, because there are some universal truths worth keeping in mind:

• No one is going to be happy with the plans that are eventually implemented because there are no “good” answers. Only less bad ones.
• Everyone is going to take a hit. Parents are going to keep shouldering heavy loads. Kids won’t get the best education. School workers are going to get shortchanged one way or another.
• There are so many variables and so many blind spots that we cannot assume that a solution which works in one place will necessarily work in another. Or a solution that works at one time will be the best option all the time.

Here are some basic facts that make figuring out school protocols extremely hard:

For starters, even talking about “school” as a singular entity is a mistake.

As an epidemiological matter, there is a world of difference between large schools and small schools, between small children and high schoolers. It makes no sense—none—to assume that the same best-practice for a room that houses a dozen five-year-olds in daycare will be the same for a high school with 3,000 students.

By the same token, we don’t know enough about the coronavirus yet to understand the differences in likelihood of, say a five-year-old catching and spreading it compared with an eighteen-year-old catching and spreading it. The data we do have on these differences is wildly conflicting.

And there are going to be large-scale geographic differences, too. On July 4, Vermont had two new cases diagnosed and Florida had 11,458 new cases. Why should we think that best practices for these two states would be the same?

Also there will be temporal differences: The true state of the infection in July may not be the case in September, which may not be the case in December.

2. Risks and needs

The one thing the data does suggest is that the risks of death from coronavirus for individuals under the age of twenty are lower than they are for people over twenty. (And the risks of death for kids under ten is lower than kids aged ten to nineteen.)

That’s obviously a good thing. But the health of the kids in school is only one part of the equation. The other is the spread of the virus. Because a kid who picks up the infection and brings it home is going to put others at risk, even if her own chances of dying are quite small.

If a family has someone over the age of fifty in the house, that’s a problem. If someone in the family is immunocompromised, that’s also a problem.

So different families will have different levels of risk.

And by the same token, different families have different needs. Just off the top of my head, here are three scenarios in which what a family needs to get out of schooling are radically different:

(1) A two-parent family in which one parent stays home and the other parent can telework has a very high level of flexibility. Maybe they can manage remote-learning with a high degree of success and a relatively small economic burden.

(2) A single mother who cannot telework and has to report to a workplace for forty hours a week. How is she supposed to manage if her kids aren't physically at school for eight hours a day?

(3) A family with a special-needs child [with significant challenges] for whom school isn’t about working towards college admissions but about early intervention to help them manage their challenges. For them, remote learning may be impossible and what the child misses from school is the critical help from a team of people who are experts in special ed.

Now overlay the different risks that families have with their different needs and you start to see, again, how it will be nearly impossible to have a one-size-fits-all protocol for opening schools.

And here I cannot stress this enough: Anyone who says either “You must open all the schools like normal” or “You cannot open any schools at all” is either foolish, or hysterical, or pushing an agenda.

You can see what I’m trying to get at here: There is no one “right” way to do school in this environment.

3. Five ideas

Mind you, I don't have the answers either. But I do have some general thoughts as to what answers will look like. Here are five broad precepts—concessions to circumstances, really—that ought to guide our thinking.

(1) Flexibility

This is the most important precept.

Every school needs to have at least three plans at the ready. One for something that looks like full opening, one for partial opening, and one for fully remote classes. If I had my druthers, I’d also want options broken out separately for elementary, middle, and high schools. And I’d want a totally separate array of options for non-mainstreamed special education.

School systems should also build in the flexibility to switch between options as circumstances warrant. If you’re in Vermont, maybe full-open works well for your district. But if your town has a spike of cases, then you have to move to another mode.

By the same token, if you’re in a Florida town where the virus is out of control in September, maybe you have to start out with full-remote schooling. But that doesn’t mean you have to stay in that mode for the entire year.

(2) Individuality

An acceptable risk for one family might be too much for another, so school systems should expend a lot of effort accommodating families. So for instance, if a school is running at partial-open and a family wants to do full-remote, there ought to be ways to make that work.

(3) An acceptance of “good enough”

Let’s just all agree, at the start, that this will be a weird school year. No one is going to get through their normal curriculum. There will be subjects that get left out or only partially covered.

And when the 2020–2021 school year comes to a close and Jimmy and Susie aren’t as far along as they should be, guess what?

The world isn’t going to end.

At the elementary level, the truth is that kids need very little other than reading and math. Middle school is largely just table-setting for high school, so if everyone is short-changed by the same amount, then the kids will be fine. And as you get into high school—particularly the higher AP levels—the kids will be more equipped to learn through textbooks and lectures. Which is what college will be like for them anyway.

(4) Pass/fail

Once you get to the high school level, a lot of institutional energy is spent trying to sift and separate kids in the 70th percentile from the kids in the 90th percentile. And even more energy is spent trying to separate the kids in the 98th from the kids in the 99th.

We could have a whole debate on whether or not this is a good thing (I vote “no”) but for the purposes of the 2020–2021 school year, we ought to just press pause on the whole thing by having everyone go to pass/fail. Grades are going to be meaningless in this environment anyway; they would all require an asterisk. Schools ought to recognize that explicitly.

“Education” isn’t a monolith. Its purposes are part childcare, part credentialing, and part genuine learning. And a lot of times that last part is the junior partner in the arrangement. By going to pass/fail, schools can put more emphasis on the learning portion this year, in whatever configuration the school year looks like.

(5) Patience

This is not going to be easy on anyone. School systems need to be patient with parents and kids. Families need to be patient with the schools. Employers need to be patient with the parents who work for them.

And we all need to be patient with each other.

1. In case you missed it over the long holiday weekend, Governor DeWine issued some preliminary guidelines for schools to reopen in the fall. (WCBE-FM, Columbus, 7/3/20) Speaking of WCBE, the future of the public radio station owned by—and now kept afloat after hundreds of thousands of dollars of fraudulently-hidden debt repayments—Columbus City Schools has still not been decided. This white elephant, still bleeding money and still riddled with old accounting problems, remains on the district’s books while someone tries to figure out something to do with it that doesn’t (apparently) include getting rid of it or (apparently) using it to help Columbus students learn anything. (Columbus Dispatch, 7/6/20)
   
   

2. Meanwhile, Columbus City Schools is moving ahead with its previous plans for a limited reopening in the fall following Governor DeWine’s new guidance. CCS, and other districts, seem to worry most about the cost of implementing whatever plan they settle on, with various different parts causing angst. I bet that Hilliard wishes it had a valuable broadcasting license to sell for big cash. (Columbus Dispatch, 7/3/20)
   
   

3. Here’s an interesting interview chatting with “I Promise” series producer Maverick Carter. There’s some interesting stuff about how the docu-series came about, and even some stuff about the actual school too. But let’s not forget that the initial academic achievement data being touted is a little old now. Time for an update! (Deadline, 7/1/20)
   
   

4. Speaking of academic achievement (well, the lack thereof), editors at Vindy.com opined this weekend that the “HB 70 experiment” has failed and CEO-style Academic Distress Commissions must be ended. Control must be returned to local voters. (Vindy.com, 7/5/20)
   
   

5. Speaking of legislation that school districts really hate, here’s another set of homeowners taking to the ballot box—as the law now allows—to have their property rezoned into a neighboring school district. This legislation must be ended, too, it seems, and control must be returned to…. Wait a minute! (Bowling Green Sentinel-Tribune, 7/6/20)

Did you know you can have every edition of Gadfly Bites sent directly to your Inbox? Subscribe by clicking here.

Thomas Sowell—writer, economist, social theorist, and currently a senior fellow at the Hoover Institution—is having a moment at age 90. He has long been a champion of high-quality education as a source of power for Black youth. His new book, Charter Schools and Their Enemies, looks at educational outcomes in more than a hundred schools in New York City. He concludes that charters, if they are able to stay true to their nature and promise, can hold the key to ending longstanding educational inequalities for Black students.

Bright spots

The Overdeck Family Foundation recently published their list of what they term “bright spots”—schools and other educational providers whose learning models adapted reasonably well to the abrupt changes required of pandemic-mitigation school building closures. A number of charter schools and networks fit that bill. The foundation was looking particularly to highlight common patterns of adaptation and common lessons learned.

Adaptation

Speaking of pandemic-related adaptations, National Heritage Academies in Michigan is working on developing curricula and lessons that work equally well in-person and online. While they and many other schools and districts were able to pivot quickly from in-person to remote learning, the variability of access to high-speed internet and connected devices remained an issue throughout the remainder of the school year. NHA reasons that even the best-laid plans for the 2020–21 school year could change unexpectedly, hence their enhanced efforts to build easily-convertible lessons to minimize disruption, downtime, and slide.

New Ohio charter schools

We learned this week that the Brilliance School is moving ahead with plans to open its new K-8 charter school in Youngstown for the fall. The Brilliance model is said to utilize a longer school day, intervention strategies to help students who are behind catch up quickly, Saturday school, social emotional learning curricula, and mentoring to help get students onto a college-prep track. They also tout a hybrid-learning component built into their model. Additionally this week, we learned even more about Cypress High School opening in Mansfield this fall. Its career-focus for students who have struggled in traditional high schools is called “a supplement to the community” by its director. “We’re not here to try to compete against any of the other high schools,” she said. “We’re all on the same team.”