15 October 2010

Poor Kids Do Better In Math At Integrated Schools

A 2002 study by the Piton Foundation, a local think tank, found that poor students in the Denver Public Schools who were in economically integrated schools did better academically than they did in schools where more than half of students were poor. The Piton Foundation study found that 53-54% of poor children who were not English language learners in the third to the fifth grade in the Denver Public Schools, at schools that had fewer than 50% poor students, were proficient in reading, while only 33% of these children at schools that had more than 75% poor students were proficient in reading. At schools where between 50% and 75% were poor, 42% of these students were proficient in reading.

A new study from Montgomery County, Maryland, which is a mostly middle class suburb of Washington D.C. with a large African-American population, came to a similar conclusion in a more rigorous natural experiment made clear that students were at integrated schools were not doing better simply because they were self-selected by parents were more motivated.

The study tracked the performance of 858 elementary students in public housing scattered across Montgomery from 2001 to 2007. About half the students ended up in schools where less than 20 percent of students qualified for subsidized meals. Most others went to schools where up to 60 percent of the students were poor and where the county had poured in extra money.

After seven years, the children in the lower-poverty schools performed 8 percentage points higher on standardized math tests than their peers attending the higher-poverty schools - even though the county had targeted them with extra resources. Students in these schools scored modestly higher on reading tests, but those results were not statistically significant. . . .

Montgomery provided an ideal laboratory because of a long-standing policy of requiring developers to set aside housing for low-income families, who win spots through a lottery. That randomness strengthens the study, researchers say. It mitigates a problem that hampered previous studies in which parents actively chose to place their children in better schools, making it difficult to separate the effect of the schools from the effect of having motivated parents.

Researchers see the results as especially significant because Montgomery . . . has been uncommonly aggressive in seeking to improve the performance of students in schools with higher poverty. It has divided the county into a high-performing, more-affluent green zone and a high-needs red zone, where schools receive about $2,000 more in per-pupil funding. And yet, the low-income students in the study performed better in the green-zone schools.

The research also supports the notion that while heredity rather than environment is the predominant determinant of IQ for children in middle class or more affluent environments, that "in poor and chaotic households, I.Q. is minimally the result of genetics — because everybody is held back. 'Bad environments suppress children’s I.Q.’s.'" Good schools help poor kids more than they help middle class and affluent kids.

Integration Addressed Only A Modest Part Of The Achievement Gap

The flip side of these findings is that the benefits to low income students of going to economically integrated schools aren't huge. The Montgomery County, Maryland study found no statistically significant improvements in reading over the seven years of the study. It took seven years for students in the more affluent study body schools to achieve an 8 percentage point improvement over their peers at high poverty schools.

This is a real and significant improvement, but it is only a moderate sized chunk of the huge achievement gap in math between well of kids and poor kids. In Colorado, for example, the achievement gap between poor students and students who are not poor in 2005 in eight grade math was 28 percentage points. If the gap in Maryland is comparable, then about a third of the achievement gap in math is attributable to the negative effects of attending high poverty schools.

Also, little of the gap in reading is related to the effects of attending high poverty schools and the gap in eight grade reading achievement in Colorado in 2005 was almost as great, 24 percentage points, and reading is probably more important than math when it comes to long term socio-economic success.

Moreover, the black-white gap in Colorado was even deeper than the poor v. not poor gap. In 8th grade math in 2005, it was about 44 percentage points. In 8th grade reading it was about 31 percentage points.

Even If Integration Is Good, How Can It Be Achieved?

Pundits commenting on the study in the Washington Post report linked above suggested that a shift in focus from spending more at schools with high levels of poverty to integrating schools is called for by the data. But, this is easier said than done.

As the Piton Foundation study explained in 2002:

Simple math suggests that Denver cannot solve this problem alone. More than two-thirds of DPS students are poor enough to qualify for free or reduced cost lunch. Because research shows that schools need to stay under the 50 percent poverty threshold, Denver cannot dilute its poverty sufficiently without either drawing private school students back into the public school system or forging an agreement with one or more suburban districts. Currently, 25 of the district’s 87 elementary schools have fewer than 50 percent of their students eligible for free or reduced-cost lunch.

This is further complicated by the fair inference that Denver's return to neighborhood schools with school choice after a desegregation order that called for mandatory busing to integrate schools was lifted is one of the main reasons that the Denver Public Schools have as many affluent students as they do. While the number of Anglo students in the system didn't immediately increase, it has grown significantly over the last decade.

Yet, the end of mandatory busing has produced a massive resegregation of the Denver Public Schools, as a Piton Foundation report in 2006 explained:

The end of busing was followed by growing concentration of minority students in intensely segregated (90-100%) minority schools. . . . In a period of two years from 1995 to 1997, during which DPS returned to a system of neighborhood schools, the share of Latino students attending intensely segregated schools more than tripled from 11 percent to 38 percent.

The increase was even more dramatic for African-American students, jumping as much as 31 percentage points during this time period. . . . In 1995, 14 percent of Anglo students attended schools that were majority Anglo. Two years later, in 1997, the percentage of DPS’ Anglo students in majority Anglo schools had more than doubled to 31 percent.

School choice within school districts, touted as a tool to reduce segregation in the Denver Public Schools, while it may have other benefits, has had a net effect of increasing, rather than decreasing the amount of racial segregation in the Denver Public Schools. A large share of students who attend schools of choice go from a home school where they would have been in a racial minority at the school to a school where they are in a racial majority or plurality.

Integration in schools in Colorado has, instead, happened at all mostly as a result of black families leaving Denver to move to the suburbs, in part because the educational choices in the Denver neighborhoods where they had been living were not good ones. As then incoming superintendent of schools Michael Bennet noted in an interview with Piton Foundation used in its 2006 study:

We know, for instance, that the primary population moving out of our schools in northeast Denver are African-American families. We want our schools to be compelling choices for everyone with school age children in Denver.

Cenus data have tracked this trend over the last decade.

The census report shows that almost all of Arapahoe County's growth since 2000 has come from the surge in the minority population. During that time, the county's white population grew less than 1 percent compared with a 70 percent growth in the Latino population and a 37 percent rise in the black population. Nine years ago, three out of four faces in Arapahoe County, home to most of Aurora, were white. Today, it is two out of three . . . .

[In its] Cherry Creek School District. . . the percentage of minority students has risen from 12 percent to 40 percent in the past decade. School officials expect half of the students to come from minority backgrounds by 2012. . . .

In Denver, the growth in the minority population slowed this decade. The black population actually fell almost 10 percent between 2000 and 2008, while the white population grew by 5 percent[.]

While the Cherry Creek Schools retain their cache, the Aurora Public Schools are struggling. The demographic makeup of the Aurora Public Schools changed dramatically in less than a decade, and the district is still trying to figure out how to best serve a new mix of students with very different needs. For example, in the Aurora Public Schools, "the percentage of English- language learners has more than doubled - from 16 percent to almost 40 percent - since 1999."

In short, integrating schools in order to improve student achievement, while still an educationally beneficial outcome when it happens, turns out to be a much harder task to accomplish than it seemed back in 1954 when the U.S. Supreme Court decided Brown v. Board of Education.

No comments: