Free lunch fraud debate still going strong

The perennial debate over fraud in the free- and reduced-priced lunch program flared once again at last night's Charlotte-Mecklenburg school board meeting. Board members have been discussing the issue for years. Back in 2008, they moved towards auditing a random sampling of applications for the federally subsidized lunch program to see how many families were cheating. But they dropped that idea after being told some $34 million in lunch subsidies could be taken away if they didn't follow federal rules stipulating that they "may do no more" than the checks prescribed by Washington. (The U.S. Department of Agriculture tells school systems to check 3 percent of applications filed by families whose yearly income falls within $100 of the cutoff levels for aid).

Anyway, the debate popped up again last night after Cindy Hobbs, head of CMS' child nutrition program, briefed the board on the latest audit results, which showed 236 applications had been checked, and 66 didn't supply verification documents matching the income reported on the application. In another 36 cases, the families didn't respond. Board members Rhonda Lennon, Kaye McGarry and Tim Morgan said if potentially 43 percent of those applications had flaws, there must be thousands of similar cases of ineligible students receiving subsidized lunches. By Morgan's math, about 11,000.

All of which prompted the same round of debate that's been going on for years on the school board. Coach Joe White, Richard McElrath and others suggested there's probably no way to find out how many people are cheating, so just feed the kids. Lennon took offense, stressing that she wasn't suggesting letting kids go hungry. She said her main concern is the fact that CMS uses the free-lunch figures to determine lots of other things, including who gets exemptions from Advanced Placement test fees and sports fees, and figuring how many additional teachers to assign schools. That prompted Superintendent Peter Gorman to repeat what he always says: the free-lunch figures, whatever their flaws, are the best tool CMS has for predicting which students will struggle and need extra help.

Lennon was undeterred.

"You're going to hear way more about this from me," she said. "That's not a threat. That's a promise."

The free-lunch fraud flap

With Charlotte-Mecklenburg Schools' poverty level creeping up and 2011 budget cuts looming, a long-running debate over the validity of school poverty numbers is ratcheting up.

CMS, like districts nationwide, uses eligibility for federal lunch subsidies to gauge school poverty (the formal term these days is "economically disadvantaged students"). The district recently released those numbers for each school. The total rose from about 51 percent of all students last year to 53 percent this year.

The cutoff for lunch aid is higher than the federal standard of poverty -- up to $40,793 for a family of four to get a reduced-price lunch, compared with $22,050 for the same-size family to be classified as living in  poverty. Eligibility is based on an honor system; with rare exceptions, CMS doesn't ask for proof of income.

Those exceptions are a small sampling that the U.S. Department of Agriculture, which administers the lunch program, requires to be verified. In 2008, the Carolina Journal, a publication of the conservative John Locke Foundation, noted that many of the families audited in CMS and other N.C. districts either failed to provide documentation or had their benefits reduced or ended based on proof of income (read that article here). The Journal talked about fraud and lying; state officials talked about family errors.

CMS's most recent sampling turned up similar results. According to Child Nutrition Director Cindy Hobbs, CMS checked 236 applications. Of those, 128, or 54 percent, maintained their benefits. Fifteen percent didn't respond, 13 percent were found not to be eligible and 18 percent were moved into a different category (most went from free to reduced-price, though a handful went the other way).

Whether you believe those numbers derive from lying, honest mistakes or a combination of the two, they do raise questions: Are thousands of families getting government aid they're not entitled to? And is CMS doling out millions of dollars in school aid based on squishy data? This year more than 74,000 students are classified as "economically disadvantaged" based on lunch applications; even 13 percent would amount to almost 10,000 who aren't really qualified.

In 2008, some board members' call for a systemwide audit of lunch applications died quickly. The USDA foots the bill -- more than $38 million for CMS students last year -- and makes the rules, and the feds don't allow that.

Many who are willing to let kids keep eating free lunches still cringe at CMS using those numbers to allot teachers (see previous blog post on that subject), supplies and other aid, especially now that the district is facing roughly $100 million in likely cuts next year.

"It's riddled with fraud," school board member Kaye McGarry said recently. "When you have millions of dollars that are allocated on those numbers, to me that is ludicrous."

McGarry agrees some schools need more help than others, and she offers no specifics on how to identify them: "It just seems there has to be a better way."

Trent Merchant, who was new to the board in 2008, initially agreed. He worked with CMS staff as they experimented with other data to identify kids who need extra help, such as the number of disabled or gifted students or those learning English. As Merchant recalls, formulas that were far more complicated yielded virtually the same results as using lunch status.

What about basing aid on the percent of students who test below grade level? After all, some middle-class kids can't read or do math, while some from low-income homes do great. But it's easy to see the downside of that: "Congratulations, Principal Smith; your staff did a great job! Your school's test scores rose so much that you'll lose three teachers next year."

The bottom line, Superintendent Peter Gorman and his staff keep saying, is that school poverty levels are the strongest, simplest predictor of academic struggles. For now, they're likely to remain the basis for aid.

But in a year filled with tough choices, don't expect this battle to end.