By the time a student finishes college, more money is spent on his or her education in America than in nearly every other country in the world.
That's because the US, compared with other developed countries, spends a lot on education. Yet all that money is yielding only middling results on international tests.
So why is American education so expensive? Partly because other social spending is low; education is expected to play a bigger role in social mobility, particularly for low-income students. And partly because education is mostly a state and local policy issue, so the way the money is spent isn't always equally distributed or particularly logical. School districts in some states spend more to educate wealthy students than poorer ones.
Court cases have forced states to divide the money more equally. But often that just increases the overall pot of money rather than redistributing it — even though the spending increases appear to make a difference in students' lives.
1) The US spends more on education but less on other social programs
That number factors in spending by everyone, not just governments. And it includes higher education — which is more expensive in the US than anywhere else in the world. Still, even when you look just at K-12, the US is spending more on each student than most other countries.
The US spends $11,193 for each student at the primary levels, more than all but three other nations — Switzerland, Norway, and Luxembourg. Those three, as well as Austria, also spend more than the US on secondary education. The US spends $12,464 per student on high school.
But nations that spend less on education are faring far better on international tests, and the US isn't seeing bigger scores as a result of its larger spending. Poland, Finland, and South Korea, where 15-year-olds performed better on those 2012 tests than American students, spend less per student than the US does.
There are a few possible explanations for this. The first is that the US spends less on social programs than some other countries. Finland spends much less per student than the US. But it spends more to reduce poverty, and across the OECD, students in poverty have lower test scores than their higher-income peers. The United States has one of the highest child poverty rates in the developed world — five times higher than Finland's. The money Finland spends to close that gap doesn't show up in the school spending numbers.
Another explanation is that US education is simply inefficient and could be better run without additional spending. Poland, for example, has made dramatic improvements in its students' performance on international tests while still spending much less per student than the US does. Even when education spending is expressed as a share of the economy, South Korea spends about as much per student as the US does and sees much better results.
2) Teachers in the US make more than teachers in other countries, but less than other American college-educated workers
About 60 percent of the $12,608 spent on each public school student in the US in the 2010-'11 school year went to instruction — paying and providing benefits to teachers and teachers' aides.
When compared with teachers in other countries, American teachers are generally well-paid: they make more at all points in their career than the average for teachers in the OECD. But teaching isn't a particularly well-paid profession anywhere. In all OECD countries, teachers make less than the average person with a bachelor's degree.
Because American salaries for people with a bachelor's degree are unusually high, that gap is wider in the US than anywhere else. In other words, teachers are well-paid by international standards for teachers. But they're underpaid by the standards of what college graduates in the US generally make.
Schools spend much more per student today than in 1970 — more than twice that amount, after adjusting for inflation — in part because they employ many more teachers than they used to. There are now about 12 students for every teacher employed by a school, down from 22 students per teacher in 1970. (Salaries have increased only slightly: after adjusting for inflation, public school teachers make about 10 percent more today than in 1970.)
Part of the reason education spending has increased is because the number of children with disabilities has grown much faster than the general population of students, and schools are now required to educate them. Special education students cost, on average, about twice as much to educate as other students. So one reason education in the US has become more expensive is that it's trying harder to serve all students — and that can come at a price.
3) More spending doesn't seem to have increased test scores in the US
The Programme for International Student Assessment, the international test used for comparison among OECD countries, has only been around since 2000. But students in the US are performing only modestly better on American standardized tests than they did in the 1970s.
At least, that's how it seems at first when you look at scores on the National Assessment of Educational Progress, a standardized test of reading and math given to a nationwide sample of students that's known as the Nation's Report Card. Here are math scores over time:
And here are reading scores over time:
But those graphs don't tell the full story. Since the 1970s, the racial composition of American school children has also changed dramatically. In 1975, 80 percent of students who took the NAEP were white; in 2008, white students made up just 58 percent of test-takers. Students of color have historically performed worse on these tests than their white peers. So if there were really no improvement in American education, and students from lower-scoring groups made up a larger share of the school-age population than in the past, you'd expect scores to go down rather than inching up.
Concealed within those unimpressive averages, in other words, is a fairly dramatic improvement for black and Latino students, and a slight improvement for white students. Since 1971, reading scores for nine-year-olds have increased only 13 points. But they've increased 25 points for Latino students and 36 points for black students. The improvement is equally dramatic in math, where scores have increased 36 points for black students and 32 points for Latino students.
So just looking at the averages can be misleading. Still, it's difficult to argue that spending more on education caused the gain in academic performance. The past four decades have seen significant policy changes, such as No Child Left Behind, as well as spending increases. And it's difficult to untangle which, if any, of those changes was the key.
4) The Great Recession led to real cuts in education spending
The trend of spending more on education every year — uninterrupted since 1996 — seems to be in decline. For the past two consecutive years, per-pupil spending in the US has fallen. In 2012, the US spent an average of $10,667 per student, a 2 percent decline, after adjusting for inflation, from 2011.
Thirty states have cut per-pupil funding since 2008, and some of the cuts have been significant: Oklahoma is spending 24 percent less per student in the 2015 fiscal year than in 2008, before the recession. At the same time, property values fell, making it harder for local districts to collect the same amount of property tax as before without raising taxes.
The budget cuts have led to teacher layoffs and larger class sizes. The federal stimulus helped shore up state and local education spending through the 2011 fiscal year, but then it expired, leaving states to cover the gaps. Although states are spending more this year than in the past, in most cases budgets still haven't recovered from the recession.
5) Rich kids have more money spent on their education than do poor kids
Historically, local property taxes have provided much of the support for education. This means wealthier areas with higher property values had more resources for their children's education. The federal government tried to close this gap through grants to schools with a high population of students in poverty as well as to schools located in areas without a lot of taxable property.
Even including federal funding, though, nationwide the poorest districts have slightly less money per pupil on average than the richest.
States are providing more support to public schools than they used to, in part to try to break the vicious cycle where high-poverty areas had fewer resources to educate their needier students. Sometimes, though, state formulas don't alleviate inequality. In Illinois, for example, Chicago public schools have more poverty than average but get less money per pupil.
This is why over the past 40 years, virtually every state been sued over how it sends money to school districts. At first, the lawsuits focused on making sure the same amount was spent to educate poor children as rich ones.
Since the 1990s, lawsuits have sought more resources for districts with large numbers of poor students, arguing that states have the responsibility to ensure students get not just an education, but an adequate education. And the cost of an adequate education for children with disabilities or who are learning English, to give two examples of students who might require more resources, is higher than for well-off students without those challenges.
Those lawsuits have driven up education spending. But they've also made more funding available to students who need more help, and research suggests the decisions made a difference.
6) School funding has gotten fairer in the past few decades — and it's made a difference in students' lives
State by state, education spending per student doesn't correlate strongly with test results: states that spend more per student, such as New Jersey and Massachusetts, tend to have strong results, but Idaho and South Dakota also get respectable scores despite spending much less. Studies in the 1960s and 1980s found no correlation between school spending and standardized test results.
But recent research argues that focusing solely on test scores misses other positive effects of school spending. A working paper published in early 2015 by the National Bureau of Economic Research looks at the long-term effect of court decisions that forced states to spend more on low-income districts.
For low-income children, more money made a big difference. A 10 percent spending increase each year in kindergarten through 12th grade, researchers found, led students to complete a few more months of school, to earn 7.25 percent more, and to be less likely to be poor. Those aren't improvements that show up in test scores, but they suggest that spending more on education made a long-term difference in students' lives.
"Money does matter and … better school resources can meaningfully improve the long-run outcomes of recently educated children," the authors, C. Kirabo Jackson, Rucker Johnson, and Claudia Persico, wrote.
7) Money matters, but how it's spent matters just as much
Jackson, Johnson, and Persico found that even when more spending made a difference, the effect was bigger for low-income students than for better-off ones. Students who weren't low-income didn't see the same effects on lifetime earnings and poverty rates.
That's one reason the researchers caution that spending alone isn't enough. The other is that simply spending more money isn't likely to make a difference unless the money is spent well.
Schools could increase spending by sending teachers on lavish retreats, which probably wouldn't make a difference. Or they could pay teachers salaries more in line with what other bachelor's degree holders make, which some people argue would impact the quality of teaching. The researchers argue that spending more on instruction and support for students was key, and that smaller class sizes, more time for instruction, slightly higher teacher salaries, and more adults in the building were probably key.
Still, students in wealthy districts in the US have more spent on their education than poorer students in 23 states. And research suggests the marginal returns on that investment are lower: compared to wealthy students in other countries, US students still underperform on standardized tests.
That suggests that money matters more for some students than for others, and that it's not just poverty holding American students back.WATCH: '10 things they don't talk about at graduation'