There are four ways to measure how much our country spends on students of color
compared to white students. Each tells a different version of the same story. First,
across the country schools spent $334 more on every white student than on every
nonwhite student.
To get this figure I simply divided each school’s adjusted total
spending into “white” and “nonwhite” shares based on the proportion of students
who are white. Then I added each of these “white” and “nonwhite” shares across the
country and divided by the total number of white and nonwhite students. This is a
nontrivial spending difference, given that the median per-pupil spending was $4,038
(see appendix for explanation of per-pupil calculation). The $334 average shortfall is
8 percent of the median per-pupil spending.
The second version of the problem is more troubling:
It focuses on the most
racially isolated schools. More than one-third of the students represented by this
new dataset attend schools that are either more than 90 percent white or more
than 90 percent nonwhite. The spending difference between these schools is large.
The mostly white schools spent $733 more per student than the mostly nonwhite
schools,9
or 18 percent of the median per-pupil spending nationwide.
How big a problem is this for students in the high-minority schools? What could
that money buy? The average-sized, mostly minority school has 605 students. This
means that the average school serving 90 percent or more students of color would
see an annual increase of more than $443,000 if it were to be brought up to the
same spending level as its almost-entirely-white sister schools. The average firstyear teacher in the United States is paid $36,780; the average teacher with 11 years
to 20 years of experience earns $47,380.10 This funding could pay the salary for 12
additional first-year teachers or nine veteran teachers.11 Alternatively, this funding
could pay for any number of other useful personnel or resources such as school
counselors, teacher coaches, or laptop computers.
OK, skeptics might say, but that is comparing the two ends of the spectrum:
schools that are almost entirely white with schools that are almost entirely nonwhite. What about the whole story? The last two versions of the problem answer
this question. Version three compares the “high-minority” schools (90 percent
or more nonwhite) to all other schools. We spend $293 less per year on students
in these heavily minority schools than on students in all other schools. That’s 7
percent of the median per-pupil spending.
Finally, across all schools, an increase of 10 percent in students of color is associated with a decrease in spending of $75 per student.
12 This is a fairly small number,
given that the median per-pupil spending in 2009 was $4,038. This analysis does
not imply that spending is being determined by race, explicitly or implicitly. But
given the 20 percent gap in high school graduation rates between white students
(78 percent) and their peers—Hispanic (58 percent), black (57 percent), and
American Indian (54 percent)—spending less money on schools that serve more
students of color, even if it is not on purpose, simply does not make sense.13
Table 2 shows how an increase of 10 percent in students of color is related to
per-pupil spending in each state. It shows that in 24 states an increase in the
concentration of students of color is associated with a decrease in dollars spent
per pupil.
These 24 states educate 63 percent of all students of color. In 13 states
the percentage of students of color is not related to a school’s per-pupil spending.
In 12 states an increase in the concentration of students of color is associated with
an increase in per-pupil spending. This positive news is tempered by the fact that
only 12 percent of students of color attend school in these states. (New Jersey was
excluded from the entire analysis because it mistakenly included federal spending
in its report instead of only state and local spending.)
State spending on unequal education
Relationship between school racial composition and dollars spent per pupil
A 10 percentage point increase in students of color is associated with . . .
A decrease in dollars per pupil
in 24 states
No significant spending
change in 13 states
An increase in dollars per pupil
in 12 states
Vermont -$762* Maine -$122 Mississippi $16*
New Hampshire -$582* DC -$117 Virginia $16**
Nebraska -$298* Wyoming -$108 Louisiana $29*
Nevada -$257* Delaware -$106 Maryland $36*
Kansas -$188* Michigan -$4 Missouri $41*
New Mexico -$179* Florida -$3 Minnesota $99*
Connecticut -$151* Indiana $2 South Carolina $118*
Iowa -$151* Tennessee $5 North Dakota $123*
Colorado -$145* Georgia $7 South Dakota $140*
West Virginia -$125* North Carolina $12 Ohio $162*
Idaho -$120* Massachusetts $16 Montana $180*
Oregon -$114* Arkansas $26 Alaska $409*
California -$104* Utah $28
New York -$104*
Wisconsin -$100*
Texas -$95*
Rhode Island -$78*
Pennsylvania -$73*
Oklahoma -$53*
Washington -$50*
Illinois -$42*
Arizona -$37**
Kentucky -$30**
Alabama -$20*
*p < 0.05, ** p < 0.10
Source: Author’s analysis of newly released U.S. Department of Education expenditure
Because a full 35 percent of students of color attend school in either California
or Texas, I highlight per-pupil spending patterns in these two states. Table 2 shows
that schools in both states have a negative relationship between the percent of
students of color and dollars spent per student. The problem is starker when we
focus on those schools serving almost only nonwhite students.
• In the California schools serving 90 percent or more nonwhite students, perpupil spending is $191 less than at all other schools, and $4,38014 less than at
schools serving 90 percent or more white students.15
• In the Texas schools serving 90 percent or more nonwhite students, per-pupil
spending is $514 less than at all other schools, and $911 less than at schools
serving 90 percent or more white students.16
Just how big are these differences? In California the average high-minority school
has 759 students. If an average-sized school got an extra $4,380 for every student, it
would mean an extra $3.3 million per year. If it were to get a more modest boost of
$191 per student to bring it in line with the majority of schools in the state, then it
would get approximately $145,000 extra per year. What could that buy? New teachers in California are paid approximately $45,000 a year, and veterans with 11 years
or more of teaching experience are paid an average of $68,000 a year.17 If per-student
funding were increased in the schools serving almost entirely students of color to
the same level as the rest of the state’s schools enjoy, it would pay the salaries of two
experienced teachers or three new teachers, or buy any number of other valuable
educational inputs such as computers, guidance counselors, or teaching coaches.
In Texas the average high-minority school is 708 students; new teachers are paid
$39,150, and veterans earn $47,110 each year.18 If an average school in the Lone
Star state were to receive an extra $514 in per-pupil funding—enough to bring
it up to the level of spending the rest of the schools in the state enjoy—it would
be able to pay the salaries of seven veteran teachers or nine new teachers.
The bottom line: Across our country, we are spending less on students of
color than on white students, at least when it comes to the state and local dollars reported by states under the new reporting requirement in the American
Recovery and Reinvestment Act of 2009.