The Education Gadfly Weekly

Last month, editors of The Youngstown Vindicator, one of Ohio’s most respected newspapers, made an unusual appeal on their op-ed page. They asked the state superintendent of public instruction, Richard Ross, to take over their local school system.

The Youngstown Board of Education had, in their opinion, “failed to provide the needed leadership to prevent the academic meltdown” occurring in their district. They added that Mr. Ross was “overly optimistic” in believing that the community could come together to develop a plan to save the district. Therefore, they pleaded, “[W]e urge state Superintendent Ross to assign the task of restructuring the Youngstown school system to his staff and not wait for community consensus.”

It’s not every day that local citizens ask the state to take charge of educating the children in their community. Such a move illustrates the despair that many Americans feel about their own schools—and their inability to do much to improve them.

That’s why, over three years ago, we at the Thomas B. Fordham Institute, along with our friends at the Center for American Progress, began a multi-year initiative designed to draw attention to the elephant in the ed-reform living room: governance. Given its ability to trample any promising education improvement—or clear the way for its implementation—it was high time to put governance at center stage of the policy conversation.

Our “anchor book” for that initiative, Education Governance for the Twenty-First Century: Overcoming the Structural Barriers to School Reform (January 2013), demonstrated how our highly fragmented, politicized, and bureaucratic system of education governance impeded school reform. One promising innovation it identified was the “recovery school district” (RSD)—an alternative to district-based governance that became a household name after Hurricane Katrina pummeled New Orleans. As new state-created entities charged with running and turning around the state’s worst schools, these districts are awarded certain authority and flexibility—such as the ability to turn schools into charters and to bypass collective bargaining agreements—that allow them to cut the red tape that has made so many schools dysfunctional in the first place.

Tennessee policymakers took note of the RSD’s success and in 2011 created the Achievement School District. Yet, outside of the Volunteer State, these alternative models have been met by policymakers and educators with way more resistance than welcome. By our count, recovery districts have been pushed in at least seven states since 2011; few have seen the light of day.

Last winter in Mississippi, house and senate bills establishing an “achievement school district” both died. The same thing happened last spring with a house bill in Texas (though gubernatorial hopeful Greg Abbott is now attempting to resuscitate it in his education platform). This summer, a Virginia circuit court judge ruled that statewide turnaround districts were unconstitutional in that state. More recently, in Georgia, Governor Nathan Deal urged lawmakers to “consider” the Louisiana model as one way to improve failing schools (it has yet to gain momentum). Likewise, officials in New Jersey and Wisconsin have toyed with the idea of statewide districts, but nothing more.

Why is it so hard for these new arrangements to gain traction? Opponents tend to complain that the districts divert funding from public schools (forgetting that they are still public) and that they remove control of schools from local oversight, handing them to state authorities and even (gasp) charter school operators.

Enter the Education Achievement Authority (EAA) in Michigan. It shares basic similarities with its brethren in Louisiana and Tennessee in that all three are charged with resuscitating the state’s worst schools within the confines of a separate, autonomous district.

But unlike the RSD in the Bayou State—which has over eighty schools statewide—the EAA is so far a more modest effort, responsible for just fifteen schools, all in Detroit, with further expansion stymied. Like the Achievement School District (ASD) in the Volunteer State, the EAA was created in response to the Race to the Top competition. Yet it is an interesting hybrid of both existing models: it combines the governance reforms of the RSD and ASD with a big push for competency-based, blended learning. And that’s what has made news: tech-oriented bloggers are singing the praises of the daring new learning platform the Authority developed, while those opposed to the whole idea of the EAA are lamenting that its students are being used as guinea pigs for market-greedy entrepreneurs.

This makes for good melodrama, but really, what are the takeaways of the EAA for other districts? To find out, we enlisted Nelson Smith, former head of the National Alliance for Public Charter Schools, who is now senior advisor to the National Association of Charter School Authorizers (NACSA). Nelson has also held senior positions at the U. S. Department of Education, the D.C. Public Charter School Board, and New American Schools. He’s keenly aware of the challenges in forging alternative educational options for kids—and in implementing recovery districts, particularly after having authored insightful reports for us on such efforts in Tennessee and Louisiana.

The EAA model—direct-run schools with limited reliance on chartering and a high-tech approach—is far from the catastrophe that some critics claim. Yet the critics aren’t all wrong. There have been many hurdles, and there is some validity to both the EAA’s claims of progress and the criticism that early results are disappointing. Some students don’t respond well to the online component or can’t handle the autonomy they’re given over their own learning. An instructional cocktail for low-achieving students that mixes competency-based, blended, and student-centered learning is tricky—and doesn’t work for all students.

In the end, the EAA was rolled out on a tight timeline. On a shoestring budget. Amid urban decline in Detroit. It would have taken a miracle for this to work out well. (Which is something policymakers might have considered before pursuing this path.) Further, its governance arrangement is a Rube Goldberg invention of epic proportions.

What’s more, officials needed adequate charter funding to woo high-quality operators to the Motor City. They didn’t have it—and they didn’t get them. And the inaugural superintendent of the EAA, John Covington, has since stepped down amid news of enrollment declines, budget woes, and other challenges.

Still, the EAA is not the complete disaster you may have heard it to be. But it’s also not a success like the RSD or ASD—both of which are improving outcomes, albeit slowly, for kids.

Which might make its cautionary lessons that much more important for other states thinking of going down this route.

The key takeaway is that neither statewide school districts nor blended, competency-based learning are silver bullets. Combining the two is a particularly precarious proposition. Furthermore, states that want to embrace this approach to school turnarounds need to create conditions that are essential to success. Michigan’s effort—though laudable, and in many ways heroic—was hobbled from the start from too many compromises and too little political support.

As with most reforms—think charter schools, or teacher evaluations—this strategy is only worth doing if done well. When it comes to educational improvement, half measures and work-arounds are rarely enough.

photo credit: erikadotnet via photopin cc

Election Day is less than a week away. Given the heat around major education policies—especially Common Core and teacher evaluations—there is increased attention to public attitudes about education. A number of polls from major news organizations, education groups, and universities have been commissioned over the past several months, and education pundits and advocates on all sides of current reform debates have endlessly parsed the results.

Unfortunately these pundits are mostly misguided, and public opinion polls on education don’t mean what people think they mean. What follows are three conclusions, all based on data from these various polls, and a discussion of what they ought to mean for education policy and advocacy going forward.

Conclusion 1: Americans’ views on education are incoherent.

The most straightforward conclusion from existing polling data is that Americans’ views are all over the map and, depending on the issue, either nuanced or contradictory. The clearest example of this is on standardized testing. The 2013 Phi Delta Kappa (PDK)/Gallup poll found that just 22 percent of the public thought that standardized tests have helped local public schools. But when asked about specific test-related policies—some of which are even more ambitious in scope than our current testing regimes—Americans express strong support. An Education Next poll, for example, shows 71 percent of Americans support mandatory high school exit exams. And despite 54 percent of respondents telling PDK/Gallup in 2014 that standardized tests aren’t helpful, between 75 and 80 percent were “very or somewhat supportive” of college entrance tests, promotion tests, and high school exit exams.

Using test scores for teacher evaluation and tenure decisions shows further contradictions. PDK/Gallup found that 61 percent oppose including standardized tests in teacher evaluations, yet Education Next found that 60 percent support requiring the demonstration of student progress on state tests as a condition for tenure. Predictably, testing opponents latched onto PDK’s results but ignored Education Next’s, suggesting that they either cherry-picked supportive research or misinterpreted the public’s true views on this issue.

In other areas, Americans appear split between “traditional” views and more “reform-friendly” ones. For instance, people generally think that we should spend more money on our public schools, a cause célèbre of the anti-reform left. Education Next found that Americans favored spending increases over decreases by a 44-to-9 percent or 60-to-7 percent margin, depending on the question. (Of course, when asked if taxes should increase to pay for higher spending—and how else would we pay for it—the margin shrinks dramatically to 26 to 18 percent.) And virtually every poll shows that Americans support more local control of education.

Voters overwhelmingly favor charter schools, a typical reform position. On the 2014 PDK poll, support for charters was between 63 and 70 percent, depending on wording. A PACE/USC Rossier poll showed 57 percent thought California should increase the number of charters. Education Next found strong support for other choice options, such as scholarship tax credits (60 percent favor, 26 percent oppose). And voters seem to be skeptical about the role of teachers’ unions, with PACE/USC Rossier finding that 49 percent of voters believed teachers’ unions have a negative impact on California education, versus 31 percent who believed the opposite.

Conclusion 2: Americans don’t pay attention to education or get involved.

If Americans’ views on education seem incoherent or contradictory, perhaps it’s because they aren’t paying attention. Take the Common Core. Despite having been adopted more than four years ago in nearly every state and covered with thousands of news articles per month for well over a year, vast swaths of the American public report not knowing about the standards. The Education Next poll showed 57 percent had never heard about the Common Core; NBC News pegged it at 47 percent. When you factor in folks who report only knowing a little about the standards, the proportion jumps to 67 percent in a University of Connecticut poll and more than 80 percent in the PACE/USC Rossier poll.

This lack of basic awareness carries over to numerous other policies. California recently revamped its educational funding plan, implementing a new Local Control Funding Formula (LCFF), the state’s most significant reform of education finance in a generation. Despite the policy having been passed several years ago, and despite requirements to involve the public in LCFF meetings, 76 percent of Californians reported having heard “nothing” or “not much” about the policy. Similarly, in 2013, 56–57 percent of voters had never heard of Race to the Top or waivers to No Child Left Behind.

The most obvious explanation for this basic lack of knowledge is that citizens don’t participate in education-related activities. PACE/USC Rossier found that less than 20 percent of voters attended school board meetings, joined a parent-teacher association, or volunteered in schools. And just 20 percent said they could definitely name a member of their school board. Voters appear to have little interest in education at the local level—even though 56 percent support greater local control.

Conclusion 3: Much of what Americans think they know about education policy is simply wrong.

Even when they claim to know something about education, alarming numbers of Americans believe things about education that are factually wrong. For instance, Education Next found that 69 percent of those who claimed to have heard about Common Core thought it meant that the federal government would collect detailed data on individual students’ test performance; another 51 percent thought that the federal government required states to use the standards. The UConn poll gave voters three options describing the Common Core, and voters selected the factually correct one at just above chance (37 percent).

Voters are also wrong about charter schools. PDK found that 57 percent of Americans think charters can charge tuition, and more than two-thirds think charters can choose students on the basis of their ability. Forty-five percent of California parents in the PACE/USC Rossier poll believed charters perform better than traditional publics (just 4 percent believed the opposite), despite that being untrue in the state. Voters are even off the mark on educational spending, consistently guessing figures that are much too low.

So, where does this all leave us? We know that voters demonstrate little knowledge of education and its major reform efforts. We also know that when they do claim to know something, they’re often factually incorrect. And yet advocates on all sides leap at the chance to construe single data points from individual polls as validating reform (or anti-reform) agendas. Clearly, none of this makes sense.

Usually, the best way to use public polls to make responsible, informed judgments about attitudes toward policies is to consider results from multiple polls, with different wording, conducted by different outfits. Yet, in the case of education, this might not work. The beliefs of the average voter probably aren’t a good guide for ed policy when a large proportion of the information underlying those beliefs is just plain wrong. Indeed, given the overwhelming evidence that American voters are uninformed and hold substantial misconceptions about numerous policies, incoherence may well be the defining characteristics of Americans’ attitudes toward our public schools. That is, it may not be that Americans’ views on education are all that nuanced after all; rather, they may reflect profound ignorance and misunderstanding of our education system and the policy efforts we have pursued over recent years.

Polls are a useful guide to what people know about education and the misconceptions they hold, and not much more. Politicians might use them to see which way the wind is blowing and how to frame issues in a way that resonates with the public. But policymakers might be advised to take poll results with a big grain of salt.

photo credit: hugovk via photopin cc

• With Election Day on Tuesday (go vote!), Education Week is covering interesting state battlegrounds with its delicious “Caravan of Delights” series. Each day, State EdWatch mainstay Andrew Ujifusa walks readers through an important state race, including polling numbers, candidates’ education-related positions, and local factors influencing the debates. Individual sections are devoted to each Republican and Democratic candidate, and a catchall section lists other things to know, such as relevant information about select superintendents and lieutenant governors. It’s worth a read for anyone who wants to make an educated vote—which we could use more of.
• Bloomberg Philanthropies is teaming up with Khan Academy, the College Board, and a handful of other prominent ed groups to boost the college-going rates of poorer students. Spending more than $10 million over the next two years, the program targets top high schoolers who come from lower-income families and aims to educate them about on the college application process. Too many of these kids are missing out on opportunities. The plan is to reach about 70,000 students, persuading at least 10,000 of them to attend any one of 250 colleges where lower-income kids tend to graduate at a high rate with manageable debt. No matter how you look at it, this is a great initiative. Still, one wonders whether it’s enough. Should interventions start younger? 

In the era of Race to the Top, waivers, and waivers of waivers, the role of state education agencies (SEAs) has increased dramatically: taking on school turnarounds, teacher-evaluation systems, and now Common Core implementation. Many argue that SEAs need "more capacity" to do these new jobs successfully. But what if we are asking departments of education not only to do too much but also to do things that they weren't built to do—and probably cannot do well? Should we shrink the SEA and its role and empower other entities to lead state-level reform instead?

Join the Thomas B. Fordham Institute for a discussion on the role of state education agencies and their leaders in the education-reform ecosystem.

PANELISTS
Deborah A. Gist - Commissioner, Rhode Island Department of Education
Mark Murphy - Secretary of Education, Delaware Department of Education
Andy Smarick - Bernard Lee Schwartz Policy Fellow, Thomas B. Fordham Institute & Partner, Bellwether Education Partners

MODERATOR
Chester E. Finn, Jr. - President, Thomas B. Fordham Institute

The testing pushback, a college boost for poor kids, adolescent readers, and school-supporting nonprofits.

Amber's Research Minute

"The Rise of School-Supporting Nonprofits," by Ashlyn Aiko Nelson and Beth Gazley, Association for Education Finance and Policy (Feburary 2014).

The case for character education hardly needs to be made. Have a glance at the motivational posters lining school hallways everywhere. “Genius is one percent inspiration, ninety nine percent perspiration,” Thomas Edison counsels our kids. “Hard work beats talent when talent doesn’t work hard,” adds NBA star Kevin Durant. Perhaps Brookings will issue a classroom poster with Richard Reeves’s face and his conclusion from this paper: “Smarts matter, but so does character.” We get it. Among the least surprising findings in social science research is that people who have certain character strengths (this paper focuses on “drive” and “prudence”) do better in life. Whether our children have great or modest gifts, we hope they will work hard, delay gratification, and persist when things don’t come easy. Still it’s easy to get nervous when Reeves suggests “too little attention is paid by policymakers to the cultivation and distribution of these character skills.” What exactly would such attention look like? Demanding that schools making AYP in grit and prudence? Character value-added measures? Likewise, eyebrows may rightfully be raised when Reeves suggests that “character skills may count for a lot – as much, perhaps, as cognitive skills – in terms of important life outcomes.” That so? A figure in the report is headlined “Drive and Prudence Matter as Much as Book Smarts for HS Graduation” (“Book Smarts?” Seriously, Brookings?), but the bar graph clearly shows “high reading skills” matter a lot more. Therein lies the mischief. It’s a lot easier to discern and measure the impact of traits like “drive” and “prudence” than to plan interventions to cultivate and grow it. And schools need no additional reasons to short-shrift academics. It’s helpful to have our faith in “performance character” issues validated, but let’s not go crazy (“teaching the whole child” is already a too-convenient excuse for poor performance). The danger of saying character matters as much or more than academics is that it might fall on fad-prone education ears as, “Academics are arbitrary. Education is all about grit!” And that wouldn’t be prudent.

SOURCE: Richard V. Reeves, Joanna Venator, and Kimberly Howard, “The Character Factor: Measures and Impact of Drive and Prudence,” The Brookings Institution (October 2014).

This new study asks a question that is receiving increasing attention: How does teacher preparation affect student achievement? To answer it, the authors gathered data from about 22,078 North Carolina educators, including how teachers were prepared and characteristics of the schools where they teach. This was combined with five years of test score data from 1.18 million students. The study is more robust than similar research, owing to its comprehensive data set and the way that it grouped teachers: Instead of lumping teachers into two broad groups—traditional or alternative certification—it creates much more nuanced groups of teachers by the way they were prepared, as well as by grade and subject taught. The first comparison is between teachers who were traditionally prepared to those who received alternative certification (meaning they didn’t have a full credential when they began teaching), excluding teachers prepared by Teach For America. Alternative entry teachers are significantly less effective (as determined by value-added measures) than traditionally prepared teachers in middle school math and high school math and science. There was no difference in the other grade levels and subjects. Second, compared to traditionally-prepared teachers, TFA teachers are more effective in six of the eight categories: elementary math and reading, middle school math, and high school math, science, and English. Third, teachers prepared out of state are less effective than those prepared in state in elementary math and reading and high school math. Fourth, teachers who began teaching with a graduate degree are less effective in middle school math and reading, but more effective in high school science. Fifth, and finally, there was no difference in any grade level or subject between in-state teachers who received their certification from private versus public schools. The study also confirms previous research that shows that there’s significant variation in teacher effectiveness within different preparation categories, not just between them. Unfortunately, it’s difficult to draw a clear conclusion about state policy based on the findings. Quality control is necessary, but creating barriers to entry into the teaching workforce can keep out some of the very people who would serve students best. Individual teacher candidates should be judged on their own merits, no matter how they are prepared (and preparation programs, regardless of their format, should be rigorous). Assessing teacher quality requires schools and districts to evaluate their merits rigorously before and after teachers enter the classroom, as opposed to assuming a teacher is good at their job simply because of how they’re prepared.

SOURCE: Gary T. Henry, et al., "Teacher Preparation Policies and Their Effects on Student Achievement," The Association for Education Finance and Policy (2014).

For the first time this year, the College Board released the annual test results of its three programs—AP, SAT, and PSAT/NMSQT—in one report. The news is mixed. On the upside, an unprecedented number of students, including a large increase in minority and low-income students, participated and succeeded. Of the 1.67 million students who took the SAT, nearly half were minorities and nearly a fourth were low-income students. And the number of high school students who succeeded on at least one AP exam (earning at least a 3 out of 5) doubled in the past year. On the other hand, the results reveal at least three problem areas. First, too many students are missing out on opportunities. Thirty-nine percent of the 684,577 students who showed AP potential (indicated by high PSAT/NMSQT scores) didn’t enroll for a single AP class. Likewise, for SAT takers, 9 percent were close to achieving the college and career readiness benchmark and might have succeeded with less than a year of additional instruction. The SAT Benchmark is a combined score of 1550 and indicates a 65 percent likelihood of achieving a B- average or higher during the first year of a four-year college. Second, using this same SAT figure, the majority of high schoolers still aren’t prepared for college or career; and the proportion who are, 42.6 percent, hasn’t increased in the last year. Third, achievement gaps abound. Only 15.8 percent of African American, 23.4 percent of Hispanic, and 33.5 percent of Native American SAT takers will be ready come freshman year, compared with 42.6 percent overall. The report offers a vivid and holistic view of today’s students. And, despite some encouraging progress, we have a long way to go.

SOURCE: “2014 College Board Program Results,” The College Board (October 2014).