Here are some facts to organize the discussion. First, here's a figure showing total number of students getting school lunch over time. The number receiving free lunches has risen substantially; the number paying for lunch has dropped.
Another angle on this same data is instead of looking at total numbers, look at the proportion of students in each category. About 60% of all students are provided a lunch at school. The share of those who are eligible to get a lunch, and actually getting one, is about 90%. The share of students who would need to pay for their own lunch, and are paying for the school lunch, is down in the last few years.
The National School Lunch program cost $11.6 billion in 2012, according to a USDA fact sheet. Why is it leading to fewer paid lunches? Perhaps the obvious explanation for fewer paid lunches is the 2007-2009 recession and its aftermath. It seems plausible that a number of families who weren't eligible for free lunches were concerned about saving some money, and started sending their children to school with a home-packed lunch instead. But this answer seems incomplete, because the program has been tweaked in a number of ways in recent years.
For example, Ralston and Newman explain:
In 2010, Congress passed the Healthy, Hunger-Free Kids Act. The Act addressed concerns about the nutritional quality of children’s diets, school meals, and competitive foods available in schools (those not part of the school meal, such as a la carte items or foods and drinks sold in vending machines). ... In implementing the Act, USDA promulgated rules requiring lunches to include minimum servings per week of specific categories of vegetables, including dark green and red/orange vegetables, as well as changes to increase whole grains while limiting calories and sodium. ... These rules took effect starting with school year 2012-13. Some school lunch standards were gradually phased in. ... The updated standards set a ceiling on total calories per average lunch in addition to existing minimum calorie requirements, with upper restrictions ranging from 650 kilocalories (kcal) for grades K-5 to 850 kcal for high schools. Total sodium levels for average lunches offered were also limited for the first time to 1,230 milligrams (mg) (grades K-5), 1,360 mg (grades 6-8), and 1,420 mg (grades 9-12) per average lunch by July 1, 2014, with intermediate and final targets scheduled for school years 2017-18 and 2022-23.
It's easy to find surveys of school administrator who cheerily praise these new rules. As someone with three children in public schools, my anecdotal evidence is that not all children are pleased with the changes. Also, I think the kinds of concerns over what children eat for lunch that motivated the passage of the 2010 act are also leading some families to believe that a home-packed lunch will feed their children better. In addition to the menu changes, the law has also led many school to raise the price for school lunches. Again, Ralston and Newman explain:
The Paid Lunch Equity provision requires districts to work towards making the revenue from paid lunches to equal the difference between the reimbursement rates for free lunches and paid lunches. For example, in school year 2014-15, the reimbursement rate for free lunches, including an additional $0.06 for compliance with updated meal standards, was $3.04 and the reimbursement for paid lunches, together with the additional 6 cents, was $0.34. The difference of $2.70 would represent the “equity” price. A district charging $2.00 for a paid lunch would be required to obtain an additional $0.70 per meal, on average, by gradually raising prices or adding non-Federal funds to make up the difference over time. Until the gap is closed, districts must increase average revenue per lunch, through prices or other non-Federal sources, by 2 percent plus the rate of inflation, with minimum increases capped at 10 cents in a given year, with exemptions under certain conditions.
Higher prices for paid lunches run the risk of reducing participation. A nationally representative study from school year 2005-06 found that a 10-percent increase in lunch price was associated with a decline of 1.5 percentage points in the participation rate of paid lunches, after controlling for other characteristics of the meal and the school foodservice operation. Another nationally representative survey conducted in 2012 found that lunch prices rose 4.2 percent in elementary schools and 3.3 percent in middle and high schools, on average, between school years 2010-11 and 2011-12. Applying the earlier results on effects of differences in lunch prices on paid-lunch participation rates, these price increases would be expected to lead to declines in participation rates of 0.6 percentage points for elementary school and 0.5 percentage points for middle and high school. These estimates suggest that price increases related to the Paid Lunch Equity provision could have contributed modestly to the decline in participation rates for paid lunches.
Other changes to the school lunch program are just beginning to be phased in. In the current school year, the "Smart Snacks in School" rules kicked in, requiring that "competitive" foods sold in schools along with school lunches "must meet limits on calories, total and saturated fat, trans-fat, sugar, and sodium and contribute to servings of healthy food groups."
After a few years of pilot programs, the eligibility rules for free school lunches are being eased. "Overall NSLP participation may also be helped by the Community Eligibility Provision (CEP), a new option that allows schools in low-income areas to offer school meals at no charge to all students. Under CEP, a district may offer all meals at no charge in any school where 40 percent or more of students are certified for free meals without an application ... An evaluation of seven early adopting States found that CEP increased student participation in NSLP by 5 percent relative to comparable schools that did not participate in CEP. The increase in overall participation associated with CEP may result not only from the expansion of free lunches, but also from reduced stigma and faster moving lunch lines due to the elimination of payments."
After a few years of pilot programs, the eligibility rules for free school lunches are being eased. "Overall NSLP participation may also be helped by the Community Eligibility Provision (CEP), a new option that allows schools in low-income areas to offer school meals at no charge to all students. Under CEP, a district may offer all meals at no charge in any school where 40 percent or more of students are certified for free meals without an application ... An evaluation of seven early adopting States found that CEP increased student participation in NSLP by 5 percent relative to comparable schools that did not participate in CEP. The increase in overall participation associated with CEP may result not only from the expansion of free lunches, but also from reduced stigma and faster moving lunch lines due to the elimination of payments."
Like a lot of middle class families, we use the school lunch program as a convenience. Our children take home-packed lunches most days, but some days the lunches never quite get made. My sense is that he nutritional value of the lunches our children take to school is considerably better than what they eat when they buy a school lunch (remembering that what they actually eat is not the same as what the school tries to serve them). But for a lot of low-income families, the school lunch program is nutritional lifeline. The poverty rate for children in the United States (21% in 2014) is considerably higher than for other age groups.
I'm sympathetic to the notion that the food served in schools should be healthier. But as a parent, I've learned that serving healthier food to children is comparatively easy. Having children eat that food is harder. And having children learn healthy habits related to food and diet can be harder, still.