For a hundred eighty days each year, fifty million kids file into a hundred thousand public schools and receive free education that costs $800 billion per year. Around half of these kids get to their school using a free bus service that costs $30 billion per year. At school, these kids receive free bathrooms, free playgrounds, and free access to gyms, textbooks, and computer equipment. If they play sports, they often receive free uniforms and free access to weight rooms and other sports equipment.

Around 90 percent of kids use the free schooling service, with the remaining 10 percent opting for a private religious school (7.5 percent) or a private nonreligious school (2.5 percent). Public school attendance is more common among kids from lower and middle income families, but the vast majority of upper income families also attend public schools.

Around the middle of each school day, the free schooling service is briefly suspended for lunch. Instead of providing free lunches to all of the students, schools charge students $0, $0.40, or approximately $4.33 for their lunch. How much each kid is charged is based on their family income except that, if a kid lives in a school or school district where 40 percent or more of the kids are eligible for free lunch, then they are also eligible for free lunch even if their family income would otherwise be too high.

Each year, schools serve 4.9 billion lunches to a monthly average of 30 million kids. Before COVID, in 2019, 68.1 percent of the kids were charged $0, 5.8 percent were charged $0.40, and 26.1 percent were charged the full $4.33. In the latest year, due to temporary COVID changes, the same numbers were 96 percent, 0.3 percent, and 3.3 percent, respectively.

The total cost of the 4.9 billion meals is around $21 billion per year. In 2019, user fees covered $5.6 billion of this cost. In 2022, user fees covered $0.7 billion of the cost. The rest was covered by public subsidies as part of the free and reduced price lunch program.

The approximately $5.6 billion of school lunch fees collected in 2019 were equal to 0.7 percent of the total cost of K-12 schooling. In order to collect these fees, each school district has to set up a school lunch payment system, often by contracting with third-party providers like Global Payments. They also have to set up a system for dealing with kids who are not enrolled in the free lunch program but who show up to school with no money in their school lunch account or in their pockets. In this scenario, schools will either have to make the kid go without lunch, give them a free lunch for the day (but not too many times), or give them a lunch while assigning their lunch account a debt.

Eligibility for the $0 and $0.40 lunches is based on income, but this does not mean that everyone with an eligible income successfully signs up for the program. As with all means-tested programs, the application of the means test not only excludes people with ineligible incomes, but also people with eligible incomes who fail to successfully navigate the red tape of the welfare bureaucracy. Research on the community eligibility program — a program that makes every kid in a school or school district eligible for free lunch so long as 40 percent of the kids are eligible for it — shows that the program improves food security and nutritional outcomes even among low-income kids who should have already been eligible for free school lunch.

The case for free school lunch is initially the same as the case for child benefits more generally. By socializing the cost of children — including through paid parental leave, child care benefits, K-12 education benefits, and through a cash child benefit — you help equalize the conditions of similarly situated families with different numbers of children. Socializing the cost of children also smooths incomes across the life cycle by ensuring that, when people have kids, their household financial situation remains mostly the same.

Indeed, this is actually the case for the welfare state as whole, not just child benefits. When people become elderly or disabled in our society, their costs are socialized and they are provided a monthly cash benefit, public health insurance, and long-term care. Children are not really any different from the elderly and the disabled. In fact, they are sort of a combination of the two: like the elderly, children’s ages make it so that they should not work, and like the disabled, children’s limited capacity makes it so that they cannot work (in fact, it’s illegal for them to work). For all three populations, worklessness makes it impossible to receive personal labor income, meaning that some other nonlabor income system is required.

Beyond this general case, there are other considerations that are unique to children and to school lunch. Well-resourced children tend to be more productive and less destructive adults, something that benefits the society overall. In the case of school lunch, well-fed kids learn better and are less likely to engage in bad behaviors that are distractive to their peers. Children also go on to become the workers of the society and thus go on to make it possible for earlier generations to retire in old age.

Conservative criticisms of getting rid of school lunch fees mostly fall into two buckets. The first is that the fees serve an important pedagogical function in society to get people to understand personal responsibility. The second is that, because the school lunch fees are means-tested, they serve an important income-redistributive function in society.

Both arguments are hard to take seriously.

Notably, conservatives don’t apply the first argument to any other part of the free schooling bundle nor to free schooling itself. The closest analogue to the school lunch service is the school bus service. The two services cost about the same, but, with an occasional oddball exception here and there, the buses are funded entirely without user fees while the lunches are funded 26.6 percent by user fees and 73.4 percent by public subsidies.

Do conservatives believe that the free bus service, which goes back to the 1800s, is destroying personal responsibility in society and that, in order to resurrect it, we need to start charging means-tested school bus fares? I’ve never seen them say it. Nor have I seen them say that any other aspect of the current free schooling bundle, including arts and sports, should charge means-tested fees.

So from the conservative discourse on this, we are apparently meant to believe that $800 billion a year of free schooling services is compatible with creating a personal responsibility ethic, but that rolling an additional $5.6 billion of spending into that service by eliminating the means-tested school lunch fees is not. This is just an obviously stupid and laughable position.

The conservative argument that means-tested school lunch fees serve an important income-redistributive function is both untrue and at odds with their general attitudes on, not just redistribution, but on how child benefit programs specifically should be structured.

In the last couple of years, we had a big public debate about whether one of the country’s cash benefits for children, the Child Tax Credit, should be extended to the poor. The conservative position on that was that child benefits should specifically be designed so that they exclude the poor and only go to the middle and upper class because such a design will increase the employment rate.

Thus, conservatives believe it is appropriate for a married family with $400,000 of income to receive $2,000 per year per child in cash benefits from the federal government, but somehow think it is wrong for that same family to receive around $800 in free school lunch. And then, on the other end of the scale, conservatives think a family with $0 of labor income should not receive a single dollar from the Child Tax Credit because that will promote worklessness but then think they should receive $800 in free school lunch benefits. Does this not also promote worklessness? It’s all very jumbled.

As far as achieving income redistribution through the application of $5.6 billion of means-tested school lunch fees goes, there are simply much better ways to go about it. As I’ve noted many times before, if you want to reduce the resources of people above a certain income by $5.6 billion, the best thing to do is apply a tax to everyone with income above that level, not dump the entire $5.6 billion charge on families that currently have children in school.

Such a tax would have a larger base and thus represent a smaller share of the income of each person taxed and such a tax would smooth incomes over time. Also, this tax-based approach would allow you to eliminate the means test, which is administratively costly and ends up excluding many low-income kids from school lunch due to administrative burdens, and allow you to eliminate the school lunch payments system, which is administratively costly and forces schools and parents to needlessly give out money to payment processing companies.

My own child currently rides the free bus to school and then, due to the community eligibility program, receives a free breakfast and lunch at the school. Today’s breakfast is an egg and cheese sandwich and a choice of a banana, apple, or orange. Lunch is a choice between cheesy pasta, a PB&J sandwich, and a turkey and ham sandwich with sides of rolls, carrots, broccoli, and fruit. I didn’t have to create a school lunch account and then input my banking information in order to load it up with money. I don’t receive any emails about the balance being low and needing to be reloaded. I don’t have to worry about it at all. Parenting is already full of stresses and hassles. Dealing with one-off payments systems for means-tested school lunch fees is one such hassle I am happy to do without.