America spends a lot more on schools than on police

And in international terms, our funding of both is very average

Since the summer of 2020, I’ve noticed a steady drip of misleading commentary that overstates the scale of law enforcement spending relative to other more worthy items, most notably education.

I think people making these assertions are typically sincerely confused. The structure of local government in the United States is very complicated; it’s not unusual for a city to exist as a separate budgetary entity from the school district that serves its population, with public safety occupying a huge share of the municipal budget alongside some desultory “youth services” item that excludes school spending. But while I do think this line of argument reflects a genuine misunderstanding, I find it frustrating that a lot of outlets give it a pass rather than attempting to clear up this confusion.

I think this has helped entrench some misunderstandings in the minds of a lot of casual observers. When I try to correct inaccurate viral tweets about school versus police spending, I hear from a lot of progressives who say something like, “the real point is we spend too much on police.”

muna @Muna_Mire

there is no money for parks. no money for arts. no money for schools. only cops. more cops. cops cops cops cops cops.

I don’t want to re-relitigate the police funding question,¹ but polluting the environment with inaccurate facts makes it harder to reason about real tradeoffs.

Schools currently receive about five times as much money as police departments, which means cutting public order spending wouldn’t generate large increases in education spending. That’s not to say spending more on education is necessarily a bad idea, but you’d first need to convince people that paying higher taxes is a good idea. That hardly seems inconceivable to me — lots of countries have higher taxes than the United States of America — but as we’ve written before, taxes are the real issue here.

Misunderstanding local education finance

In the United States of America, both education and policing are primarily local functions, albeit ones that have state and federal money sloshing around.

Local government in the United States is very convoluted and also varies a lot from place to place. But outside of New England, it’s fairly common for people to live outside the boundaries of any incorporated municipality. My in-laws, for example, live in Kerr County, Texas just outside the town limits of Ingram and very close to the larger town of Kerrville. I often say they live in Kerrville because that’s the name of a town that people familiar with Texas will recognize, but they don’t actually live inside the boundaries of any town at all. The upside to not being inside the Ingram borders is you don’t need to pay taxes to the Ingram municipality. The downside is you don’t get Ingram’s services, including the services of the Ingram Police Department.

Children in this country have the right to attend public school, so you can’t just say “well, if you live in an unincorporated part of the county, you don’t get any education.”

This could be resolved by making education a county government function, and some places do it that way. But Kerr County doesn’t. Instead, the territory is divided up into school districts that are based in different towns and generally named after them. So there’s an Ingram Independent School District that runs an elementary school, a middle school, and the high school. The elementary school serves kids from the town but also kids from nearby unincorporated areas. A bit to the west, the Hunt Independent School District runs a separate elementary school for kids who live near Hunt. But those kids shift over to the Ingram ISD for middle school and high school. The point is that while these ISDs loosely correspond to the towns, they are separate tax and budget entities.

So when Alexandria Ocasio-Cortez says that “40% of Uvalde’s city budget goes to police,” she is absolutely right. But that amounts to approximately $4 million per year, whereas the Uvalde Consolidated Independent School District spends around $41 million per year. Most of that is state money, but the local taxing entity provides UCISD with $12 million per year. It’s difficult to do a genuine apples-to-apples comparison because UCISD serves a larger territory than the Uvalde Police Department. And an accurate understanding of the budget certainly doesn’t prove that Uvalde’s police expenditures are useful (their police department seems terrible!), and their school system may have unmet funding needs.

Knowing the scale is still important, though. If the police budget were much larger than the school budget, then a modest trim in police spending could generate a large increase in school spending. But even the dramatic step of halving the police department would facilitate only a minor rise in school spending. Uvalde clearly needs to tackle some serious issues with its local law enforcement, but nothing they do there is going to be a game-changer for local education finance.

America spends more on schools than cops

If you ignore the complications of local government structure, the byzantine flows of state and federal grants, and look just at the national aggregate, you can see that the United States spends a lot more on K-12 schools than it does on police. If you add in prisons and courts, the gap narrows a little — especially after the huge incarceration boom of the 1990s — but it’s still a large gap.

The United States also spends about 0.9 percent of GDP on direct public financing of tertiary education, which is roughly equal to national spending on police (and that ignores the de facto subsidy of the federal student loan program).

What’s more, since the baby boomers went to college, K-12 spending as a share of GDP has been roughly flat at about 3.5 percent, even though school-age children have been falling as a share of the population. So if you plot school spending per student versus police spending per capita, you see a larger and growing gap.

Now, again, maybe we have the balance wrong. Some conservatives are very critical of soaring per-student spending on K-12 schools in a way that strikes me as misguided. Both policing and education are labor-intensive activities, and their costs tend to rise over time for Baumol’s Cost Disease reasons. Where possible, we should be trying to inject productivity-increasing technology like automated traffic enforcement into these fields, but the opportunities are not always there. Schools specifically used to benefit from a huge implicit subsidy derived from women’s sexist exclusion from most forms of professional work. There was never a moment when the United States of America officially decided to adopt feminism and then explicitly thought through the second-order consequences of that for the public school system. But it’s clearly had a big impact in terms of what it costs to hire and retain great people — perhaps a post for another day.

Just as a budget fact, though, police are a small line item compared to schools. And this puts us broadly in line with global norms.

America’s school/cop priorities are pretty normal

The United States of America is an international outlier in a number of very real ways. Private spending plays a uniquely large role in our health care system, and the same is true of our higher education system. Unusual doesn’t mean bad — I’m a defender of our higher education finance setup — but it does raise the question.

And in terms of crime and crime control, the United States is a clear outlier on two dimensions: the number of people locked in prison and the number of guns circulating on our streets. A social phenomenon like “young men form gangs that fight each other” has drastically different implications in a society when those fights are conducted with guns versus knives and bludgeons.² This is at least part of the reason why our police are more heavily armed and trigger-happy than European cops, and you might think it means we fund policing more generously than other rich countries.

But in reality, this does not seem to be the case; OECD data indicates that the United States spends a pretty average amount on policing.

On education spending, we are the median member of the G7, spending more on our K-12 schools than Japan, Germany, or Italy but less than Canada, France, or the UK. The OECD average is to spend 3.1 percent of GDP on K-12 education, and we spend very slightly higher than that at 3.2 percent of GDP.

It’s also worth noting that there isn’t a clear tradeoff here in the international data.

Japan and Germany both spend less than we do on police, but they also spend less on schools. France and the UK spend more than we do on schools and also more on police.

Of course there are additional factors you could consider. We have more kids than Japan does, so we probably should spend a larger share of GDP on schools for them. We also have way more guns floating around, so we probably should spend a larger share of GDP on police than Japan does. I think most progressives would say the real solution here would be to have fewer guns, and I basically agree. But in the short-term, getting rid of guns not only poses daunting political problems but would also require a surge of policing to enforce new rules.

Budgetary tradeoffs are hard

Last summer, Adam Serwer suggested a national campaign to disempower police unions and reduce cops’ job protection.

That idea never really secured a lot of progressive movement support because even though it doesn’t literally implicate them, it’s uncomfortable for the public sector unions that (unlike police unions) serve as integral pillars of progressive coalitions. By contrast, labor unions representing teachers and other municipal workers are very comfortable with the idea of shifting budgetary resources away from cops and toward other local government functions. And that’s all fair enough — stakeholders in local fiscal politics are bound to fight with each other and throw some elbows in the process.

Personally, I support a larger overall public sector that would install air conditioners and air filters in public schools while expanding Medicaid and investing in policing.

But that means being willing to pay higher taxes as well as trying to persuade people of the merits of higher taxes. In the city where I live, I would really like to convince people that they should be open to rezoning for the construction of much more market-rate housing, in part because this would have large fiscal benefits that make it easier to do this stuff.

In a world where you can’t convince people to pay higher taxes, public officials are forced into sharp tradeoffs with no easy answer, and police departments and schools really are in competition for funds. There are surely some jurisdictions whose funding balance is tilted too far one way or the other. But in terms of the longer-term project of securing support for public investment, it’s very damaging to persuade people with progressive instincts that we could make all our dreams come true by raiding some gargantuan police budget. The truth is we already spend a lot more on schools than on cops, and our levels of spending on both are very normal in international terms, even though some aspects of the fundamentals suggest we should have above-average spending in both areas.

1. Here’s the case that police funding is effective at reducing crime, that reducing misconduct will require higher rather than lower budgets, and that crime control should be seen as a meaningful social justice priority.

2. This is why Stephen Spielberg and Tony Kushner can brilliantly remake “West Side Story,” but you can’t really update it to a modern setting — if the gangs had guns from the beginning, the whole story is very different. But you could absolutely do a contemporary West Side Story set in England.