Decoding Congress’ Stealth Formula for Raising the Price of School Lunch

In its recent re-authorization of the school meals program, Congress included a provision that would force schools to raise the price they charge students who don’t qualify as low-income.

Some hailed this little-noted mandate as a way to generate what the USDA estimates could be a $2.6 billion windfall for schools over 10 years. But I wondered, did this estimate include all the kids who will stop buying the much-maligned school lunch if it gets more expensive? And how could the federal government possibly know how many kids would drop out of the lunch line rather than pay the higher tab?

It took some weeks of prodding, but I finally obtained an internal document from the USDA’s Food and Nutrition Services branch revealing how the agency at the Senate’s behest formulated the magic number--$2.6 billion—that it then passed on to the committee where the price mandate was hatched.

It also confirms what I originally suspected: The government can’t really know what will happen to participation in the lunch program if schools raise prices. You might expect the federal government to bring every possible resource to bear and weigh ever so carefully a decision that stands to affect some 90,000 public schools. But that is not the case.

“We do not prepare or publish cost estimate memoranda in the way that the Congressional Budget Office sends such materials to members of Congress,” explained a USDA spokesman. “There was no formal document or methodological write-up given to the committee. The agency gave the committee a number in answer to a question.”

So here’s where the number came from:

School food experts for years have known that raising the price of lunch means some students who pay “full price”--about 12 million of the 32 million who participate in the school lunch program on any given day—will stop buying it.

In 2005, Mathematica Policy Research, contracted by the USDA to conduct one of the agency’s periodic reviews [PDF] of the school meals program, created a statistical model that estimated 56 percent of paying students would buy lunch if the price were $1.50, but that fewer—50 percent—would pay if the price were $2.00.

It’s important to note that Mathematica did not conduct a study of children’s actual purchasing behavior at different price levels. What they gave the USDA was a modeled prediction based on all sorts of data the firm collected from 2,314 students at 398 schools that year, including the types of food served, the amount of time kids were given to eat, prices charged, and interviews with children and their parents revealing what the kids typically ate in the course of a day and family income.

Based on Mathematica’s prediction within this narrow price range, Food and Nutrition Services extrapolated its own formula in order to respond to the Senate committee’s request for an estimate: For every cent the price of lunch increases, students who pay full price will drop out at a rate of .11 percent. It then calculated that the Senate’s proposed lunch price mandate would generate $2.6 billion more income over 10 years—and cause nearly 500,000 paying students to stop buying lunch.

But that’s hardly the end of the story. Under the new mandate, schools will be required to raise prices each year by an amount equal to the rate of inflation plus two percent until they are completely caught up with what the USDA estimates is the actual cost of providing a school lunch, currently $2.72. Many schools now charge as little as $1.50.

As if things couldn’t get any more complicated, the government’s baseline is a moving target. What the USDA calculates as the cost of providing lunch—the amount it gives schools to pay for a fully-reimbursable meal--is adjusted upward

annually with the rate of inflation. Thus, the vast majority of schools will take longer than 10 years to reach the government’s baseline. Nearly half will take more than 20 years.

Could the USDA’s formula for calculating drop-outs possibly hold up that long and under all sorts of different economic conditions?

“They [the USDA] asked us that, and we told them we had a problem with it,” said Mathematica senior researcher Anne Gordon, one of the report’s primary authors. “I don’t remember exactly, but I think it was around $3 we couldn’t make a prediction. We can’t know what will happen at that price, because none of the schools we looked at charged that much.”

In other words, accepting the USDA’s predictions years into the future requires a leap of faith. “It’s probably the best they can do,” Gordon said.

The prospect of annual price hikes out to the horizon has caused great alarm among the nation’s lunch ladies. In the current recession, they are grappling with millions of dollars’ worth of meals eaten by children whose families are deemed able to pay, but haven’t.

School food service directors opposed a congressional edict to raise prices, but would have preferred a House version that “sunsetted” the law after 10 years and required the USDA to conduct an impact assessment after four years. But in a last-minute rush to enact the child nutrition legislation, that version never came up for a vote.

When I asked the School Nutrition Association, representing some 53,000 school food workers, to comment on the new law, they reported results from some recent price increases.

When the lunch price rose 15 cents to $1.75 in Munster, Ind., in 2008, for instance, nine percent of the kids dropped out. In Caroline County, Md., the price rose 35 cents to $2 in 2007 and participation plummeted 16 percent. In Franklin Township, Ind., schools hiked the price 15 cents to $2.10 in 2009 and 12 percent of the kids stopped buying. Schools in Willoughby-Eastlake, Ohio, raised the price 10 cents in 2008 to $2.60 and participation fell 10 percent.

SNA spokeswoman Diane Pratt-Heavner said that while food service directors accept Mathematica’s 2005 report as “the most comprehensive data available to FNS, they question it’s accuracy in portraying how families will react to current price increases.”

“The economy is worse than in school year 2004-2005, and their own experience tells them that participation drops when you increase prices,” Pratt-Heavner said.

The School Nutrition Association is asking the USDA to test the lunch price increases on a pilot basis before imposing the mandate nationwide.

Besides higher prices, other looming factors will likely suppress school lunch participation and upset the cafeteria business model. Upgraded nutrition standards—including more helpings of vegetables, more whole grains, fewer French fries and other potato products, and much less salt in food—are expected to cause more paying kids to reject the federally-subsidized hot meal.

In a 2010 report to the USDA [PDF], Mathematica predicted that adopting a full range of improvements to make meals “healthier” would result in 5 percent of elementary school children dropping out of the program, and even more—12 percent—at the secondary school level.

Sociologist Janet Poppendieck, whose book on the federally-subsidized meals program, Free for All, has become a widely-cited text, rejects the idea of forcing schools to raise prices across the board.

Undercharging may give an unfair advantage to some families who can afford to pay at the expense of low-income children. But Poppendieck says the USDA is probably underestimating the number of parents who will react angrily to higher prices and pull their children out of the lunch line. And that could hurt the entire program’s ability to function.

She fears for millions of children on the margins--those who aren’t exactly affluent, but don’t qualify as “low income” either.

“If we lose them, it’s not just the loss of children, we lose the claim that this is not just a welfare program,” Poppendieck said. “And the more school lunch has the label of being a welfare program, it imposes a kind of a shame tax on kids who do want to participate. I think that’s the wrong direction to go.”