Your Schools

Squishy numbers: Why it’s harder than ever to talk about poverty in CMS

Eligibility for subsidized school lunches used to serve as a measure of student poverty. Now that has gotten complicated.
Eligibility for subsidized school lunches used to serve as a measure of student poverty. Now that has gotten complicated. Courtesy of Charlotte-Mecklenburg Schools

As student poverty emerges as a central issue in the future of Charlotte-Mecklenburg Schools, it’s getting harder to talk about.

I don’t mean because “poverty” often serves as code for a tangle of issues related to race, social class, opportunity, cultural values and neighborhood identity. That’s true but not new.

The latest challenge is more basic. Poverty numbers that have long served as the basis for big decisions in CMS, from teacher assignments to millions of dollars in aid, have gotten squishy.

Normally the nuances of education data matter mostly to a handful of educators, reporters and policy geeks. But the school board has embarked on a review of student assignment. Many board members and community leaders say that reducing concentrations of poverty is the best way to give all kids a fair shot at a promising future.

In other words, those numbers could eventually shape where your kids go to school, what your home is worth and whether Mecklenburg County can compete for high-paying jobs.

My big fear is that people are throwing around numbers that aren’t accurate. That can really harm public schools.

Amy Hawn Nelson of the UNC Charlotte Urban Institute

Yet if you look for up-to-date reports on school poverty you won’t find them. That’s because CMS currently has two different ways of tallying those levels, depending on the school. There’s a good reason, but it’s complex and confusing – exactly the kind of thing that undermines trust among people who are inclined toward skepticism – so CMS never posted its 2014-15 poverty report.

“My big fear is that people are throwing around numbers that aren’t accurate,” says Amy Hawn Nelson, director of social research for the UNC Charlotte Urban Institute “That can really harm public schools.”

Nelson recently co-authored a book advocating for CMS to use socioeconomic status to promote diversity in schools. Now she’s worried that confusing poverty data, misunderstood or misused, will undermine the effort.

So if you care about public education, join me in a walk down Wonk Lane to understand the challenge.

It’s about school lunches

In the wake of the Jim Crow era, courts forced CMS and other districts across America to integrate public schools by race.

In the 1990s and early 2000s, as courts overturned race-based assignment, people began to talk about school diversity in terms of socioeconomic status. Some districts, such as Wake County, used it to shape student assignment.

In CMS it became a minor factor in a complex plan that mixed geographic assignment and school choice. As schools came to look more like the neighborhoods that surround them, some saw poverty levels soar. When that happened, many of the remaining middle-class and affluent families fled.

In an effort to compensate for the increased challenges, CMS has pumped in millions of dollars and assigned additional teachers to keep classes smaller at high-poverty schools.

The old system may have been controversial, but it was consistent

School poverty levels, here and nationwide, are based on the percent of students who qualify for free or discounted school lunches, a federal program based on family income.

There have always been questions about those numbers. The income standards for lunch aid are higher than federal poverty levels – $44,863 vs. $23,850 for a family of four, for instance. And because there’s little or no verification, some conservatives charge that the levels are inflated by freeloading families and/or school leaders seeking extra resources.

It may have been controversial, but the system was consistent.

Until last year.

The numbers get wacky

That’s when CMS started using a relatively new federal option designed to eliminate paperwork and make sure students get the nutrition they need to focus on learning.

At schools where most students had consistently qualified for aid, CMS shifted to providing free breakfast and lunch for everyone. That means families don’t have to fill out forms every year and school employees don’t have to process them.

It also means CMS can no longer rely on lunch status to track poverty at those schools.

The district can tally students who qualify for food stamps and other forms of public assistance. But that yields a lower number.

Under a new federal formula, some CMS schools have poverty levels higher than 100 percent

Here’s where your face may start to twitch: To compensate for the difference, the federal government spells out that the public-assistance number be multiplied by 1.6 to create a school poverty rate for schools using the free-meals-for-all approach (OK, you know there’s a cumbersome government label: It’s Community Eligibility Provision, or CEP).

For some CMS schools, that creates poverty rates higher than 100 percent.

Perhaps you can see why Scott McCully, the CMS official in charge of reporting these numbers, simply gave the board a list of 2013-14 poverty rates as it began its student assignment review. He’s trying to figure out what to do this year, when the rates are tallied in October. He says he may end up releasing dual reports: One with numbers for the 74 CEP schools (presumably with the 100+ percentages adjusted down) and another with old-fashioned school lunch data for the remaining 94.

Why it matters

In the short term, the inconsistencies mean CMS is grappling with how to handle teacher allocations and other formulas based on free-lunch rates.

In the longer term, the challenge of dealing with concentrated poverty led two candidates for Charlotte mayor to talk about capping school poverty levels as Wake County used to do.

At last week’s school board meeting, member Rhonda Lennon tried to explain why higher poverty levels here complicate that issue. She cited the CMS rate as about 61 percent, a significantly higher number than I’d heard. Two days later, McCully told the board 56 percent was his best estimate for 2014, with 2015 numbers yet to come.

You could argue that the details don’t matter, whether it’s 56 percent vs. 61 percent in CMS or 93 percent vs. 100 percent at any given school. Anyone who’s paying attention can identify schools that have overwhelming levels of poverty and all the challenges that come with that.

But decisions about individual schools and students need to be based on solid data. And if CMS rolls out a new and confusing set of numbers, some folks are bound to question the accuracy and the motives.

Nelson, the Urban Institute analyst, says districtwide numbers matter, too. “We’re at a point in the community where if we inflate the number unintentionally, that can drive people out of our public schools,” she said last week.

There are other ways to talk about poverty, diversity and schools, Nelson says, and she contends it’s worth the work to come to a new consensus on reliable data.

Add that to the checklist of challenges ahead.

  Comments