feeling the heat

Higher temperatures equal lower test scores — study confirms that students learn less in overheated classrooms

PHOTO: THE DENVER POST PHOTO BY HYOUNG CHANG
Darlene Cisneros , 10, left, and Manuela Ramirez , 10, of Harry M. Barrett Elementary School, Denver, Colo, is taking a note by the electric fan during the afternoon writing class on Tuesday afternoon in 2003.

A warm classroom is not conducive to learning, as any student trying to pay attention to a teacher’s lecture on a hot day can attest. That’s not lost on teachers.

“I really dread school based on the weather, especially in the spring and in the fall,” said one teacher in Baltimore in a school without air conditioning. “If it’s really hot … certainly [student] engagement goes down.”

Now, there’s research to back that up.

A new study, released through National Bureau of Economic Research on Monday, shows that after a particularly hot year of school, high schoolers performed worse on the PSAT, an exam taken to prepare for the SAT and determine winners of the National Merit Scholarship.

“Hotter school days in the year prior to the test reduce learning, with extreme heat being particularly damaging and larger effects for low income and minority students,” write the paper’s four researchers. “On average, a 1 degree Fahrenheit hotter school year prior to the exam lowers scores by … slightly less than 1 percent of a year’s worth of learning.”

The research highlights how external factors can impact students’ performance on high-stakes tests, while also suggesting that air conditioning, still missing in many schools, is a worthwhile investment.

The study, which has not been formally peer-reviewed, relies on extensive data: PSAT data of 10 million students from the high school classes of 2001 to 2014.

Paper authors Joshua Goodman of Harvard, Michael Hurwitz of the College Board, Jisung Park of UCLA, and Jonathan Smith of Georgia State focus on whether students learn less, as measured by the PSAT, in school years with more hot days. (The College Board administers the PSAT.) To get at that, they look at students who took the test multiple times, and then, accounting for the fact that students generally perform better after taking the test again, they see if students tended to do worse when the exam was preceded by a warmer year.

Indeed that’s exactly what they find, with every degree increase in average temperature above 60 degrees during the school year, leading to slightly lower PSAT scores. More days with extreme heat — over 90 or 100 degrees — also caused score drops. Impacts were significantly larger for black and Hispanic students and those in lower-income areas.

Why might that be?

“Wealthier students may be able to compensate for lost learning time by getting additional instruction from their parents or private tutors,” the authors say. “Such students may also be more likely to attend schools where teachers have better capacity to compensate for lost learning time by adjusting lesson plans or adding more instructional time.”

Heat during the summer, weekends, and holidays didn’t impact test scores, which is consistent with the idea that learning in school drove the findings.

The study then turns to the question of whether air conditioning prevented the negative effects of heat on learning. They find, that in fact, it generally did, with most of harmful consequences of heat disappearing in schools that appear to have air conditioning.

The paper doesn’t have perfect data on whether schools actually have and use air conditioning, but instead relies on surveys of counselors and students. A substantial number of students — about 42 percent — said that on hot days classrooms sometimes or frequently got too hot, though counselors were less likely to say this was an issue. Paradoxically, in hotter areas of the country, hot classrooms were less of a problem, likely because air conditioning was move prevalent.

Share of students by school district who said hot days led to hot classrooms. Source: “Heat and Learning”

Black and Hispanic students and those in low-income areas were a few percentage points less likely to have air-conditioned classroom than white or affluent students in similar climates. This may explain why those students saw steeper test score declines as the result of warm weather.

Because of this and the fact that black and Hispanic students tend to live in places with higher temperatures, the paper estimates that the impact of heat in schools explained somewhere from 1 percent to 13 percent of the racial test score gap on the PSAT.

The analysis is in line with other research on the topic, including a study of New York City showing that high school students do worse on end-of-year exams in years with higher temperature and on warmer testing days. (The latest paper doesn’t focus on the single testing-day effect because the PSAT was taken in October, when heat is less likely to be a concern.)

A recent analysis found that most of the country’s 50 largest school districts report having air conditioning in every classroom, but also that 11 districts have some or many classrooms without it.

Concerns about heat in school may have prompted some policymakers to promise action: New York City schools have vowed to install air conditioning in all classrooms by 2022.

It may well be a worthy investment, according to the latest study. “The benefits of school air conditioning likely outweigh the costs in most of the U.S., particularly given future predicted climate change,” the authors write.

Unintended consequences

When Denver stopped lunch-shaming, debt from unpaid meals skyrocketed

PHOTO: David Buffington | Getty Images

After the Denver schools chief made a high-profile announcement last August guaranteeing a full meal to students whether or not they had the money to pay, many advocates cheered the end of so-called “lunch-shaming” in the 92,000-student district.

Then came an unpleasant surprise: Debt from unpaid lunches soared, rising to $356,000 from $13,000 the year before.

Denver’s exploding meal debt — amounting to roughly 900 unpaid lunches every school day of the year — illustrates the balancing act districts nationwide face amid growing public support for policies prohibiting lunch-shaming. Such shaming often involves giving students who can’t pay small, alternative meals, putting stickers or stamps on them to remind their parents to pay, or even throwing out their meals.

In the last couple years, a growing number of districts nationwide have established policies to curb lunch-shaming. Some states, including New York, Iowa, and New Mexico, have passed statewide legislation with the same goals. The idea behind such measures is to free students from the burden of debt they have no power to pay and ensure they don’t go hungry at school. But with school districts obligated to pay for the meals, food service leaders are often left scrambling to cover mounting costs.

The school lunch debt is one reason Denver district officials quietly introduced snacks such as Doritos and Rice Krispies Treats in elementary school cafeteria lines late this past winter. The new additions, seen as unhealthy by some parents, helped generate around $41,000 in new revenue for the nutrition services department.

Diane Pratt-Heavner, director of media relations for the School Nutrition Association, said she hasn’t yet heard of another district with a debt increase the size of Denver’s following the introduction of a lunch-shaming prevention policy. But she said it’s an issue the group, which represents school food service employees, plans to watch closely.

“In many districts, allowing all kids to automatically get a free meal …. can turn into a real financial challenge for the program,” she said, noting that it can take away the incentive for parents to fill out the free and reduced-price meal application.

Nearly one-third of the district’s lunch debt last year came from families who were eligible for free or reduced-price meals, but signed up part-way into the school year, after their children had already received free school lunches. The federal government covers lunch costs for students eligible for free lunches and part of the cost for students who qualify for reduced-price lunches. For elementary school students in Colorado (and starting next year for middle-schoolers), the state covers the remaining cost of reduced-price lunches.

Another 68 percent of Denver families with unpaid meal debt don’t qualify for free or reduced-price meals. Still, district officials said it’s impossible to determine how many of those families would qualify for subsidized lunches if they applied, how many struggle financially but just miss the cut-off for eligibility, and how many can afford to pay for school lunches but choose not to.

Theresa Peña, regional coordinator for outreach and engagement in Denver’s nutrition services department, supports the district’s new lunch-shaming prevention policy, which ended the practice of giving students with lunch debt cheese sandwiches or graham crackers and milk.

Still, district officials didn’t expect the ballooning lunch debt, which at one point was projected to hit a half-million dollars, she said.

Peña said the district is stepping up efforts to get every family to fill out the free- and reduced-price meal application for next year — an extra challenge in the current political climate in which some immigrant families fear leaving a paper trail.

Last year, in addition to adding new revenue-generating snacks in elementary schools, the district tried to recoup the debt by making weekly robocalls to parents, working with principals to do outreach to families, and in some cases sending letters home with students.

“We made a pretty hard push,” Peña said. “It did make an impact, but not as great an impact as we had hoped.”

A national problem

Most districts nationwide accrue some debt for unpaid meals.

A 2016 survey by the School Nutrition Association found that three-quarters of school districts rack up unpaid meal debt, up slightly from 71 percent two years before.

In Denver, the amount of lunch debt ranges widely by school, with some accruing less than $50 and others accruing thousands. Omar D. Blair Charter School had the highest lunch debt among Denver schools last year at $11,500. Meanwhile, Florida Pitt Waller, Joe Shoemaker Elementary, Thomas Jefferson High School, and Cheltenham Elementary all reported lunch debts between $2,500 and $5,000.

At Shoemaker, where two-thirds of students are eligible for free or reduced-price meals, Kitchen Manager Chris Juarez said he believes much of the school’s $4,000 in lunch debt came from families who would have qualified for subsidized lunches but didn’t apply. Sometimes, he said, families don’t realize they have to re-submit their application each year; this fall, he plans to meet with returning families — in addition to new families — to emphasize that fact.

Other parents don’t realize they have to add to the form if a related child joins their household, he said. And language barriers may still be a problem, even though the form is available in many languages. In addition, some may worry that filling out the form means their immigration status can be tracked. A 2017 Denver school board resolution specified that the district does not collect or maintain any information on students’ immigration status.

Juarez suspects only a small percentage of Shoemaker families can afford to pay for their children’s lunches, but choose not to.

Shoemaker Principal Christine Fleming, said her top priority is making sure kids get to eat lunch, no matter what. She sees non-payment as a “parent issue,” and said, “I don’t want 5-, 6-, 7- year olds to carry that burden.”

Fleming said she’s always reserved some money in a special “principal’s account” to cover the cost of unpaid lunches, including in 2017–18, when she set aside a few hundred dollars.

Previously, that practice was common across the district, Peña said, but once the lunch-shaming policy took effect, “a lot of them said, ‘Zero out my principal account. I’m not going to do that anymore.’”

In 2016–17, when the district’s lunch debt was just $13,000, an online fundraising campaign and a contribution from a private donor covered the outstanding balance. But not this year.

A district grant of $100,000 paid off lunch debt from students who were eventually eligible for free or reduced-price lunch last school year but whose parents may not have signed up right away. Peña said the district has not finalized how the remaining $256,000 will be paid, and has until June 30 to make a decision.

Is it junk food?

Before this year, elementary schools in Denver sold some snacks — officially called a la carte items — in their cafeterias. These included turkey sticks, granola bars, popcorn, string cheese, and yogurt.

Peña said the district decided to add more a la carte items in February, a few months after district food service supervisors visited nearby districts, including Jeffco and Cherry Creek, and learned that “a la carte sales were a big deal” there.

The additions include more than a half-dozen kinds of chips, Rice Krispies Treats, gummy fruit snacks, and pistachios. All of the items — some of which are slightly reformulated versions of the same products sold on grocery store shelves — adhere to federal rules governing school snacks. Parents were not informed of the new snack offerings when they were introduced.

Susan Scovell, who has two children at Bradley International School in southeast Denver and works part-time as a personal chef, said of the new snacks, “It’s pretty much total junk food.”

She got wind of them when her second-grade daughter began mentioning that friends routinely bought Doritos and Cheetos at lunch time.

“It took me months to figure out this was going on,” she said. “Most parents really had no idea.”

Scovell said the new snacks stand in stark contrast to the district’s efforts to emphasize scratch cooking and other kinds of healthy eating initiatives, such as the week-long fruit- and vegetable-tasting event at Bradley this spring.

Peña, who said the district plans to communicate better about the snack options this coming year, said parents can prevent their children from buying certain snacks. To do so, they need to contact the school’s kitchen manager and request that a note be added to the student’s school meal account citing the restriction. She conceded that the process may not be obvious or easy for all parents, and said the department will look to address that.

Peña also said that principals or kitchen managers have the option to limit the sale of a la carte snacks at their schools. For example, they can choose not to sell certain items, or restrict the sale of a la carte items to the last 15 minutes of the lunch period or to certain days of the week.

Denver is hardly unique in offering a la carte snacks at elementary schools.

Other large Colorado districts, including Douglas County, Jeffco, and Cherry Creek, also offer such items to grade school students. All three districts allow parents to limit or block their children’s snack purchases.

Carol Muller, state director of Colorado Action for Healthy Kids, which promotes nutrition and exercise initiatives in schools, said one of the top concerns she hears from parents across Colorado is about a la carte snacks. At the same time, she understands the financial pressures school cafeterias are under.

“It’s a really tough issue for everyone involved, including us,” she said. “We certainly support food service staff. We don’t want to add a bigger burden to them, but on the other hand, as a parent, I don’t find all the snacks acceptable either.”

survey says

More bullying reported at New York City schools, study shows

PHOTO: Anthony Lanzilote

More New York City students say there is bullying in their schools, a report released Monday showed. The findings also revealed that many schools reporting the greatest number of violent incidents on campus have no social workers on staff.

The report was commissioned by New York City Comptroller Scott Stringer.

Stringer also released an audit of how school safety matters are recorded, and concluded that the education department should provide more oversight and streamline incident reporting rules.

“The audit found clear breakdowns in communication in the reporting and tracking of incidents and actions taken,” according to a press release from Stringer’s office.

The education department disputed some of the comptroller’s findings, and in a written statement, spokeswoman Miranda Barbot wrote: “We have detailed protocols in place to ensure allegations of bullying are immediately reported, investigated and addressed, and are investing in both anti-bullying initiatives and mental health supports.”

But the pair of reports raises scrutiny of Mayor Bill de Blasio’s school discipline reforms, which favor  “restorative” practices that emphasize mediation over punishment, and make it harder to suspend students.

Advocates of the de Blasio reforms say the shift is necessary because black and Hispanic students are more likely to be arrested or disciplined at school. Research has shown such disciplinary action can lead to higher dropout rates. Critics of the reforms, meanwhile, say the changes have created more chaotic schools.

The findings are also likely to add to a chorus of parents and elected officials who say more emotional supports are needed for the city’s most vulnerable students. Students who experience a mental health crisis during the school day may be handcuffed and shuttled to hospitals. The city’s latest budget, which was approved last week, includes an additional $2 million to hire social workers and guidance counselors in schools that currently don’t have any.

Here are some highlights from the reports.

More students report there is bullying in their schools — but the data comes with a catch.

Last year, the education department’s annual survey showed that 82 percent of students said their peers “harass, bully, or intimidate others in school.” That’s up year over year, and up significantly from 65 percent of students in 2012, which was the lowest rate recorded since at least 2010. (De Blasio’s discipline reforms started to take effect around 2015.)

A note about these numbers: Prior to 2017, the survey asked whether students harass, bully or intimidate other students none, some, most, or all of the time. The most recent survey responses were slightly different: none of the time, rarely, some of the time, or most of the time — a change that may have artificially inflated the bullying numbers.

That’s enough to render the survey data unreliable said Max Eden, a researcher who has studied school climate for the conservative-leaning Manhattan Institute — a critic of the mayor’s discipline reforms. Still, taken with other findings, it’s reasonable to think that bullying is on the rise at city schools, he said.

Among the other evidence: A first-of-its-kind report, released this month under a new city law, that showed substantiated bullying incidents are on track to increase this year.

Schools that log the most violent incidents often lack mental health supports.

Guidance counselors and social workers are key when it comes to creating safe schools because they can help address the root cause of violent or troublesome behavior, advocates who want more mental health supports say.

But many of the city’s neediest schools go without that help.

Of the schools reporting the most violent incidents on campus, 36 percent lack a full-time social worker, the comptroller found. On campuses where there are social workers, caseloads are a staggering 700 to one. That far exceeds the recommended ratio from the National Association of Social Workers of 250 general education students per social worker — and it’s higher than the citywide average of 612 students per social worker, according to the comptroller.

The comptroller’ compares that to the ratio of New York Police Department school safety agents who are placed in schools: There is one safety agent per 228 students, according to the report.

“Our city is failing to meet the social and emotional needs of our students,” Councilman Mark Treyger, of Brooklyn, who has pushed the city to report more up-to-date bullying data and to hire more school counselors, said in an emailed statement.

Schools may be underreporting violent incidents, something the education department disputes.

In a separate audit, the comptroller compared logs kept by school safety agents to incident reports filed by school leaders. In 21 percent of cases, incidents that were noted by safety agents were not reflected in the school reports.

The school data, in turn, are used to report incidents to the state for its Violent and Disruptive Incident Report, or VADIR. The discrepancy could raise questions about the already-controversial reporting system. (VADIR has been criticized for classifying schoolyard incidents as serious offenses, and the state has tweaked its definitions in response to those kinds of concerns.)

This finding also comes with some caveats. The comptroller looked at only 10 schools — a tiny sample of the city’s portfolio of about 1,800. And the education department took issue with the methodology.

In its response to the audit, education department officials said that the police data doesn’t align with the state’s reporting categories, and that the information may not be comparable because of student privacy concerns and recordkeeping issues on campuses where multiple schools share a building.