The Education Gadfly Weekly

This is not the new normal.

“Remote learning” will not be the primary way most American children are educated for very long, certainly not any time soon. Neither have we transformed ourselves into a nation of homeschoolers or “unschoolers” any more than passengers thrown from a sinking ship into lifeboats can be said to have taken up rowing.

Kids will go back to brick-and-mortar schools—yes, even “government schools”—at the earliest possible moment. Most of them want to go. Their parents mostly want it even more strongly. The act of sending our kids every morning to a place called a school is a cultural habit formed over many generations. It persists because we value it, not for want of a better idea or a more efficient delivery mechanism for education. Except at the margins, such as the 4 percent or so of U.S. kids who were already being home-schooled, COVID-19 changes none of that. The old normal will be back, and in some places sooner than we think. So let’s think about what that will look like, and whether we will be ready.

Start with the obvious: To throw all or even most of our efforts into remote learning is “shoe bomber” planning, responding to the last attack instead of anticipating the next one. Twenty-one states have already closed schools for the remainder of the year. It is a fantasy to believe that we can stem the effects of months without real school by ginning up instructional capacity on the fly in unfamiliar forms in the midst of a public health crisis. By all means, distribute devices and attack the digital divide. Signal to apprehensive students and parents that education must go on, keep kids attached, and strive for normalcy. Schools that have found ways to continue high-value instruction deserve attention and praise. But let’s not gull ourselves into thinking this is some sort of durable solution. It’s an emergency response, nothing more.

The availability of online learning, whether organized by schools or independently by engaged and motivated parents, will mitigate some of the worst effects of months without real school. But as Paul von Hippel of the University of Texas at Austin noted, parents “are even more unequal than technology.” The child spending these weeks sequestered at home with two college-educated parents with flexible work schedules and the bandwidth to supervise and contribute to learning will get more help and enrichment than a single parent with three children and a high-school education. “We’ve known since the 1966 Coleman Report that families are much more unequal than schools,” von Hippel wrote in Education Next. “We’re about to see what happens when we turn up the volume on families and turn it down on schools.”

What will happen is foreseeable. A report last week from the Northwest Evaluation Association’s (NWEA) Collaborative for Student Growth Research Center estimates that students will return to school in Fall 2020 with “roughly 70 percent of the learning gains in reading relative to a typical school year.” In math, the COVID-19 slide will be even greater. School children are likely to return with “less than 50 percent of the learning gains and in some grades, nearly a full year behind what we would observe in normal conditions,” according to Beth Tarasawa and Megan Kuhfeld, the report’s authors.

Local conditions will dictate local responses, but a few common themes seem certain. The youngest children are the most likely to be seriously harmed by extended school closures. Kindergarten through second grade are the most critical years in a child’s formal education, and very young children are least likely to get any value from remote learning without an active and engaged parent or caregiver working directly and extensively with them. Their social development will also suffer. As one veteran early childhood educator put it, “First grade is going to feel a lot like kindergarten.”

The first priority for forward-thinking district superintendents planning for the resumption of school is staffing. “If I was in a district, I would be asking, ‘Who has their teaching credential in the district office and can teach for the first six weeks of next year?’” says Chris Minnich, CEO of NWEA. “Anyone who has the expertise should be in the classroom because making sure that kids, especially in the early grades, get targeted instruction in the beginning of the year is essential.”

At least at the elementary and middle school levels, where the effects of lost learning time will be acute, “we have to think differently about the first six to eight weeks of school,” Minnich adds. “I’d let the experts tell us how long that should go on for, but it needs to be a fairly focused period of time, primarily focused on the previous grade standards.”

Candice McQueen, the former commissioner of the Tennessee Department of Education who now leads the National Institute for Excellence in Teaching (NIET) expects one of three possibilities for next year, depending on the district. “We’re back in brick and mortar; we are fully virtual for some period of time; or we’re in some hybrid approach. Leaders need to take a blank sheet of paper now and say, ‘What would we do in each of these scenarios?’” she counsels. Echoing Minnich, she stresses human capital. “Let’s staff and hire differently for what we think we will need. And let’s give people the right job description, training, and support for what we need them to do.”

Some district superintendents are already thinking ahead. In the first days of the crisis, Clark County (Nevada) Superintendent Jesus Jara’s first priority, as with many of his counterparts, was to “get the students engaged. Obviously take care of their social, emotional [needs].” But with digital devices in the hands of fewer than two-thirds of the district’s 325,000 students, Jara’s attention turned almost immediately to preparing for school to resume. Digital learning is “enrichment only.” De-emphasizing what can be done to advance learning right now have brought howls of derision in some quarters, particularly when the perception exists that districts would prefer no learning to inequitable instruction. But the decision is more sensible and defensible when considering how little can be reasonably accomplished under emergency circumstances versus the prodigious foreseeable challenges that lie ahead.

“I said to the team, ‘Guys, we’ve got to get ready.’ We’ve got to keep everybody calm, but our focus is moving into what is school going to look like when we return,” Jara told me. The gaps are going to be “larger for our students of color, our free and reduced lunch, our English language learners, our special ed students,” and for the 7 percent of Native American children in the district, the nation’s fifth largest. “We have to accelerate and start planning the return.”

Plans to make up for lost time require urgency, but should avoid complexity and stay well within the talents and capacity of existing staff. Adding complexity to a teacher’s job and stealing time away from the urgent task of remediation should be set aside. One idea worth considering is stepped-up use of assessment-driven ability grouping, at least in the younger grades and for the concentrated period of time suggested by Minnich. This would allow for targeted instruction with minimal differentiation. The district’s strongest teachers should be in front of the students who have fallen the furthest behind. Top-down district control of curriculum and planning may also be necessary for a time, so that classroom teachers can focus exclusively on assessment and remediation. “Teacher leaders and master teachers will be key in design and implementation,” McQueen says.

While this is no time for “moonshots” and grandiose schemes, with massive unemployment looming and a terrible job market for new college grads, there may be an opportunity to press non-professionals into emergency service for several weeks or months of targeted, high-dosage tutoring in high-needs schools.

Keep it simple. Keep it focused, intense, achievable, and time-limited. The most attention should be on those who have fallen the furthest behind. If we aren’t planning for the resumption of schools, and for the foreseeable conditions we will face, we will be caught flat-footed a second time.

Amid the plague that surrounds us, essential attention is properly getting paid to the education challenges of out-of-school kids: What can their parents, their schools, and their districts do to compensate for missed classroom time and the learning loss that’s bound to occur between now and the resumption of something resembling normalcy.

Within that universal concern, there’s been special focus on the educational needs of children with disabilities and those who were already lagging before their schools closed. All necessary, all important, all good.

Yet the plague also highlights—albeit indirectly—another set of students that’s so far receiving no special attention that I can spot: gifted and talented youngsters and their need for acceleration, enrichment and advancement designed to make the most of their abilities.

What’s the connection? Look at the population of adults we’re counting on to get us out of this mess: scientists, doctors, bio-statisticians, microbiologists, demographers, inventors, industrial leaders and economists, not to mention government leaders with the intellect and training (and discipline) to see beyond their immediate political interests.

Where are tomorrow’s Dr. Faucis to come from? The infectious disease expert that The New Yorker dubbed “America’s Doctor” grew up in southwest Brooklyn, then attended a top-flight, selective-admission (but free) Jesuit high school, which propelled him on to Holy Cross and Cornell Medical School, where he was first in his class.

But it’s not just Tony Fauci. What about all the researchers now hard at work on better tests, vaccines, and therapies? How about the brilliant team at the Institute for Health Metrics and Evaluation, our foremost source of the intricate “modelling” that enables everyone else to plan around forecasts—and constantly-revised forecasts—of what lies ahead? (You can glimpse the “credentials” of those folks here.) Not to mention the next generation of medical practitioners such as those now struggling to save thousands of lives?

Nor is it just COVID-19. It’s also the next pandemic. Climate change. Bold new technologies. Inventive social policies. Fantastic musical and literary works. And much else, including—while I’m at it—far better strategies for educating children. 

Isn’t it obvious that the more of these folks we have the better? And the more diverse they are the better? Yet to oversimplify just a bit, it’s today’s smart kids from every sort of background who are by far the strongest candidates to play those roles tomorrow. But will they—enough of them, from across enough of the demographic and socioeconomic boards—be well-prepared to succeed in those roles with the levels of expertise, knowledge, and skills to generate the breakthroughs that we’ll need?

That depends in large measure on how well we educate them today—while they’re out of school—and next year when (we hope) they’re back in school, and the year after that, and on into the future.

It may be counterintuitive, but the present situation actually poses an unusual opportunity to serve these kids better than many have been served in regular classrooms. They could be learning more right now when they’re not twiddling their cognitive thumbs sitting through lessons pitched too low for them and enduring painful efforts by regular teachers to differentiate instruction in ways that benefit them as well as the strugglers.

If America’s fate thirty years from now ends up relying on the small population of mostly-privileged smart kids whose parents can get them into elite public and private schools and supply auxiliary programs and opportunities that challenge and fulfill them intellectually, we won’t have nearly enough of the experts that we’ll need—and those we have won’t much “look like America.” 

So let’s keep all those other smart out-of-school kids in the forefront of today’s education concerns, too. Or you can be sure we won’t be ready for what’s coming.

With the COVID-19 crisis upon us, many of the rhythms of regular life have been paused. But not for high school seniors, most of whom are still making plans to head to college in the fall. And along with azaleas and allergies, early April still brings news about acceptances and rejections from selective schools.

And if you know someone with a stellar academic record, loads of extra-curriculars, and sky-high SAT or ACT scores who nevertheless didn’t get into his or her number-one pick, you are not alone. The conventional wisdom is that it’s now much harder to be accepted into highly selective colleges and universities than it was a generation ago. But is it true?

To find out, earlier this year we identified the median SAT scores (math plus verbal) of most of the one hundred top national universities and fifty top liberal arts colleges, as determined by the 2020 U.S. News and World Report college rankings, for the incoming freshmen classes of 1985 and 2016. The 1985 data came from the 1986 edition of Barron’s Profiles of American Colleges, tracked down at the Library of Congress. The 2016 data were found in the 2018 edition of Barron’s Profiles of American Colleges. Since the SAT was re-normed in 1995 and again in 2016, but after the time period for our data, we used a concordance table published by the College Board to adjust the 1985 scores accordingly. Of the 150 institutions examined, median SAT scores were available in both years for ninety-five of them.

We used SAT scores as our measure of selectivity rather than acceptance rates because those rates have plummeted for most institutions in recent years, due to changes in the application process itself. For example, the “common app” makes it relatively easy and affordable for students to apply to larger numbers of schools. More and more students, particularly those gunning for the big-name institutions, are doing so. Given that the college ranking lists give a lot of weight to rejection rates, colleges face strong incentives to push those rates as high as possible by doing things such as generating large numbers of applications from students unlikely to be accepted, much less attend. In contrast, median SAT scores give us a better measure of the academic quality of an institution’s freshman class.

To be sure, SAT scores aren’t perfect. They don’t, for example, account for the many students who took and submitted ACT scores over this time period. The scores themselves can be inflated by test prep, which is much more common now—which can give an advantage to more affluent students—or by various machinations of the colleges, such as encouraging otherwise well qualified applicants with lower test scores to show up for the spring semester of freshman year rather than in the fall. The schools in our sample look at much more than test scores, so “selectivity” goes far beyond the SAT. All that said, they do tell us something about the difficulty of getting into a given college today versus a generation ago.

What did we learn? For once, the conventional wisdom has it about right. With just a few exceptions, median SAT scores for the most selective institutions have risen significantly over the past generation, meaning, by that measure, it really is harder to get into such schools today than thirty years ago.

Among the institutions in our sample, the average median SAT score for the incoming freshman class increased by 93 points, to 1309 from 1216.

Fourteen institutions saw increases of 150 points or more:

• Elon College, to 1192 from 952, an increase of 240 points
• University of Chicago, to 1520 from 1331, an increase of 189 points
• Ohio State, to 1280 from 1100, an increase of 180 points
• Pitzer College, to 1293 from 1116, an increase of 177 points
• UCLA, to 1320 from 1143, an increase of 177 points
• UC-San Diego, to 1337 from 1160, an increase of 177 points
• Boston College, to 1360 from 1190, an increase of 170 points
• SUNY at Stony Brook, to 1290 from 1120, an increase of 170 points
• University of Pittsburgh, to 1280 from 1112, an increase of 168 points
• Georgia Tech, to 1410 from 1243, an increase of 167 points
• Case Western, to 1410 from 1247, an increase of 163 points
• Brown University, to 1490 from 1335, an increase of 155 points
• UNC-Chapel Hill, to 1263 from 1110, an increase of 153 points
• Notre Dame, to 1440 from 1290, an increase of 150 points

(View these data for all institutions in our sample here.)

Another way to understand this shift is hold median SAT scores constant, instead of holding institutions constant. For example:

• NYU’s median SAT scores today are roughly the same as Columbia’s median SAT scores a generation ago.
• Georgia Tech and Case Western have median SAT scores today that are roughly the same as Yale’s a generation ago.
• Tulane’s median SAT scores today are one point higher than Duke’s a generation ago.

Other SAT score “twins” include:

• UC-San Diego (today) and Georgetown (then)
• University of Richmond (today) and Carleton College (then)
• University of Miami (today) and Northwestern (then)
• UNC-Chapel Hill (today) and Notre Dame (then)
• SUNY at Stony Brook (today) and Middlebury College (then)
• Virginia Tech (today) and Wake Forest (then)
• Penn State (today) and Boston College (then)

(View more “twins” here.)

What might explain these trends?

First, the U.S. population has increased almost 50 percent over this time. But places like Yale aren’t admitting 50 percent more students than they were in 1980. Plus, there are more international students fighting for these slots as well.

Second, America is a richer and better-educated country than it was a generation ago. The upper-middle-class and above has grown significantly as a share of the population, and these are the families most likely to aspire to highly selective institutions for their children. And given the cost of attending these top-tier schools, they are also the families who can most afford them.

Third, as Caroline Hoxby found in a 2009 study, much of this stems from the increased mobility of America’s top students. Simply put, students with high SAT scores are more likely to travel to highly selective institutions than they were in the past, when a large proportion of high-scoring youngsters would have stayed close to home and attended less selective regional colleges and universities. As a result, such students are self-sorting into fewer, and more selective, schools.

So what are the implications?

First, parents and students should keep these trends in mind, especially if they just got bad news from the admissions offices of highly selective colleges and universities. It really is harder to get into a highly selective school today, and families should set their expectations accordingly. Even if junior is every bit as smart and hardworking as dad, his beloved alma mater may be out of reach.

Second, to the extent that employers use college selectivity as an indication of applicant quality (not that they should, but many do), they need to increase their list of colleges and universities by which they are impressed. For example, if they used to recruit from Duke, they should start recruiting from Tulane as well. (The same might apply to singles who are “shopping” for a partner from a prestigious school!)

Finally, if there’s anything the current COVID-19 crisis is teaching us, it’s that we should keep everything in perspective. Even if students got disappointing news in their in-boxes over the past few weeks, they and their families should remember that America is blessed with hundreds of world-class colleges and universities. Now let’s hope that they can open their doors again come September.

Editor’s note: This article was first published by Education Next. Mike Petrilli also discussed it with Martin West, Education Next’s editor-in-chief, on the EdNext Podcast.

The vast majority of voucher program studies have shown positive competitive effects, meaning that students who remain in public schools benefit as their schools are exposed to competition from private-school-choice programs. The latest working paper from David Figlio, Cassandra Hart, and Krzysztof Karbownik adds to the literature by examining one of the nation’s oldest and most expansive of these programs, the Florida Tax Credit Scholarship (FTCS). Its long history and rich data make it possible for the authors to provide more evidence of positive competitive effects and to attempt to define and describe the mechanisms behind competitive pressure.

Figlio and team merge student-level school records with state birth records and use a student fixed effect design to explore impacts of competitive pressure on public school students’ educational (test scores) and behavioral (absenteeism and suspensions) outcomes. The sample size is large—nearly 1.2 million students in Florida public schools between 2002 and 2017—but also limited in a few important ways. It includes only students in grades three through eight (to take advantage of consistent testing data conducted in those grades) and who were also born in Florida and attended public schools consistently throughout this time period. Students with severe disabilities were excluded. Despite this, nearly 81 percent of children in the birth record data are ultimately represented in the school data.

Consistent with the majority of previous studies, the analysts find that the FTCS program produced statistically significant benefits for students whose public schools were exposed to increasing competition from private schools. The greater the exposure, the greater the benefits. To measure the level of exposure, the analysts have developed a competitive pressure index that combines five elements of the private school landscape within a five-mile radius of any given public school: the number of private schools serving the same grade range, the distance between each public and private school, the diversity of private school types (Catholic, Lutheran, non-denominational, secular, etc.), the number of seats in those private schools, and the number of churches. It is through this mechanism, and a couple of other nuances of the research design, that they are able to isolate the effects of competition. And by varying the weights of the components, Figlio and team find that private school density appears to be the main driver of competitive pressure. In other words, benefits observed are modestly larger for students attending public schools that have more extant private competitors nearby prior to the expansion of the program. This is a new wrinkle in the competitive effects literature and deserves to be explored more fully.

Observed benefits overall include higher standardized test scores (a 10 percent increase in FTCS participation is associated with a 0.4 percent of a standard deviation greater improvement in public school students’ math and reading scores) and lower absenteeism and suspension rates among public school students. The former declines by 0.5 percent and the latter by 0.4 percent. Effects are particularly pronounced for lower-income public school students, as per previous research, but are also positive for more affluent children. While these seem like modest effects when rendered in this way, they are seen over a huge number of students and over a fifteen-year time span. That adds up to significant benefits for kids. Additionally, the widespread nature of the benefits indicates that generalized school improvements, rather than those targeted solely at scholarship-eligible students, explain the results. (For a more thorough discussion of the observed effects and the researchers’ efforts to isolate them, please check out the Education Gadfly Show podcast from February 26.)

Private-school competition, as research continues to show, improves education generally. Defining what effective “competition” really consists of in the non-market education space—and figuring out how policymakers can leverage that knowledge to benefit kids—could be a game-changer in raising the bar for all students.

SOURCE: David Figlio, Cassandra Hart, and Krzysztof Karbownik, “Effects of Scaling Up Private School Choice Programs on Public School Students,” Northwestern Institute for Policy Research Working Papers (February 2020).

Work-based learning (WBL) refers to career preparation and training that occurs within a job setting, connects to classroom and academic experiences, and involves supervision and mentoring. Recent federal efforts like the Community College Apprenticeships initiative and state-driven programs like Pathways in Delaware have increased awareness of WBL and solidified it as a critical method for preparing students for high-wage, high-demand jobs. In general, there are four main models: apprenticeships, internships, cooperative education (co-ops), and practica and clinical experiences.

Despite recent expansion efforts, WBL programs in most places still lack clear frameworks for measuring student participation and outcomes. A new report from the Urban Institute sheds some much needed light on the WBL landscape and on-the-ground implementation efforts. The analysis focuses on community colleges, as they enroll millions of students in credit-bearing courses and noncredit workforce programs, have diverse student populations, and provide career-focused degree and certificate programs that require close collaboration with employers.

To understand the broader WBL picture, the authors examined existing research and national data sources such as the Adult Training and Education Survey and the Registered Apprenticeship Partners Information System. They also conducted interviews with national experts and state agencies. Overall, the data indicate that WBL opportunities are growing in number. Unfortunately, data also indicate that disparities in access, participation, and outcomes exist across WBL models for traditionally underrepresented groups such as women and people of color.

At the community college level, there are very few nationally representative data sources that focus on WBL. Research on outcomes—such as persistence toward degrees and certifications, post-WBL employment outcomes, retention rates, and wages—is limited. Registered apprenticeships have a stronger evidence base because they are tracked by the federal government. Results show that enrollees have higher earnings. In the ninth year after program enrollment, they earned on average almost $6,000 more a year than they would have if they hadn’t participated in an apprenticeship. Results for co-op programs, which alternate classroom learning with learning on the job, are varied. Some studies show they lead to impacts such as reduced time to finding employment and higher starting salaries, while others suggest these impacts fade over time. Internships, meanwhile, have very little research associated with their impact despite being extremely popular. One study found that longer-term, paid internship programs were associated with better outcomes than short-term, unpaid programs.

To better understand how WBL works on the ground, the authors interviewed representatives from six community colleges about their experiences implementing and measuring it. By and large, these schools offer WBL on a continuum that ranges from low-involvement career awareness activities to more intensive programs aimed at career preparation and training. Opportunities are usually linked to college credit and academic or instructional components. For example, one college offers an online WBL course that requires students to submit writing assignments throughout their experience. Several colleges have dedicated staff, such as a WBL coordinator, who is responsible for managing relationships with employers.

In terms of implementation, a persistent concern for all respondents was the structure of WBL for students who have full- and part-time jobs. If a student’s job is related to their program of study, colleges reported that they attempted to make accommodations to align the job with WBL recognized by the college. When that wasn’t possible, they attempted to be flexible in other ways, such as allowing students to complete their WBL experience over a longer period of time.

Unfortunately, gathering employment and outcome data seems to be mostly aspirational for these colleges. Some are able to access state wage records, but others are limited by privacy laws and other roadblocks. They typically rely on course codes to track how many students participate in WBL each semester. Tracking WBL outside of course credits is far more difficult, though many colleges are actively working to improve their data collection efforts. Colleges also rely on student and employer feedback surveys to gauge program quality and outcomes. Low response rates and survey fatigue among both students and employers was cited as a challenge to obtaining consistent data from these efforts. Nearly all respondents mentioned that limited funding is a serious barrier to robust data collection.

The report closes with recommendations. Both federal and state governments should develop a common definition of WBL and identify data elements that should be tracked. States would be wise to incorporate WBL into their longitudinal data systems. Community colleges should consider incentivizing employers and students to complete surveys to get consistent and reliable data. And philanthropies should provide funding that can build capacity among government agencies, colleges, and employers to help track outcomes. Overall, identifying strategies to assess participation, progress toward diversity and equity goals, and outcomes is vitally important.

SOURCE: Shayne Spaulding, Ian Hecker, Emily Bramhall, “Expanding and Improving Work-Based Learning in Community Colleges,” Urban Institute (March 2020).

On this week’s podcast, Diane Tavenner, co-founder and CEO of Summit Public Schools, joins Mike Petrilli and David Griffith to offer advice on how schools can help meet students’ individual needs when they return after lengthy closures. On the Research Minute, Amber Northern examines whether social-emotional learning is improving students’ self-efficacy, self-management, and social awareness.

Amber's Research Minute

Martin R. West, Libby Pier, Hans Fricke, Heather Hough, Susanna Loeb, Robert H. Meyer, & Andrew B. Rice, “Trends in Student Social-Emotional Learning: Evidence From the First Large-Scale Panel Student Survey,” Educational Evaluation and Policy Analysis (March 23, 2020).

The Education Gadfly Show is available on Google Podcasts, Libsyn, Spotify, iHeartRadio, Apple Podcasts, TuneIn, The Podcast App, PodBean, Podbay, Chartable, Castbox, and myTuner Radio.