U.S. students from educated families lag in international tests.
Hanushek, Eric A. ; Peterson, Paul E. ; Woessmann, Ludger 等
"The big picture of U.S. performance on the 2012 Program for
International Student Assessment (PISA) is straightforward and stark: It
is a picture of educational stagnation. ... Fifteen-year-olds in the
U.S. today are average in science and reading literacy, and below
average in mathematics, compared to their counterparts in [other
industrialized] countries."
U.S. secretary of education Arne Duncan spoke these grim words on
the bleak December day in late 2013 when the international tests in
math, science, and literacy were released. No less disconcerting was the
secretary's warning that the nation's educational problems are
not limited to certain groups or specific places. The "educational
challenge in America is not just about poor kids in poor
neighborhoods," he said. "It's about many kids in many
neighborhoods. The [test] results underscore that educational
shortcomings in the United States are not just the problems of other
people's children."
It's not just about kids in poor neighborhoods
In making his comments, Secretary Duncan challenged those who cling
to an old belief that the nation's educational challenges are
confined to its inner cities. Most affluent Americans remain optimistic
about the schools in their local community. In 2011, Education Next
asked a representative sample to evaluate both the nation's schools
and those in their own community. The affluent were especially dubious
about the nation's schools--only 15 percent conceded them an A or a
B. Yet 54 percent gave their local schools one of the two top ratings.
Public opinion is split on how well the nation's schools
educate students of different abilities. In 2013 Education Next asked
the public whether local schools did a good job of teaching talented
students. Seventy-three percent said the local schools did
"somewhat" or "extremely" well at the task, as
compared to only 45 percent who thought that was true of their capacity
to teach the less-talented.
[ILLUSTRATION OMITTED]
To see whether this optimistic assessment of the nation's
ability to teach the more able student is correct, we draw upon the
latest tests of student achievement and find that, as Secretary Duncan
has said, the nation's "educational shortcomings" are not
just the problems of the other person's child. We have given
special attention to math performance because math appears to be the
subject in which accomplishment is particularly significant for both an
individual's and a country s economic well-being.
When viewed from a global perspective, U.S. schools seem to do as
badly teaching those from better-educated families as they do teaching
those from less well educated families. Overall, the U.S. proficiency
rate in math (35 percent) places the country at the 27th rank among the
34 OECD countries that participated in the Program for International
Student Assessment (PISA). That ranking is somewhat lower for students
from advantaged backgrounds (28th) than for those from disadvantaged
ones (20th).
There are examples of excellence. The six states with high
proficiency rates (58 to 62 percent) among students from families with
high levels of parental education rank among the OECD top 13 on this
measure. But students from these states are a small portion of the U.S.
student population, and other states rank much lower down the
international list. In many places, students from highly educated
families are performing well below the OECD average for similarly
advantaged students.
There can be little doubt that education shortcomings in the United
States spread well beyond the corridors of the inner city or the
confines of low-income neighborhoods where many parents lack a high
school diploma. While bright spots can be identified--particularly in
some states along the country's northern tier--the overall picture
is distressing to those concerned about the potential evolution of
economic well-being of the United States in the 21st century.
Conventional Wisdom
Not everyone agrees that the nation's schools are in trouble.
In their apology for the American school, David Berliner and Gene Glass
seek to reassure Americans by trying to isolate the problem to minority
groups or those of low income. "In the United States, if we looked
only at the students who attend schools where child poverty rates are
under 10 percent, we would rank as the number one country in the
world," they write.
These claims are highly misleading. The important question to ask
is, Do students with similar family background do better in the United
States than in other countries?
Defenders of the American school also like to compare the
highest-performing states within the United States to all students in
other countries. "Massachusetts ... scored so high that only a few
Asian countries beat it," Berliner and Glass declare. "The
states of Massachusetts, Minnesota, and Colorado ... ranked among the
top-performing nations in the world." It is true that Massachusetts
schools stand up to world competition, but it is important to keep in
mind that the K-12 students living in Massachusetts are just 2 percent
of the nation's total. One cannot generalize to the country as a
whole from this small state.
The Study
Our state-by-state data come from the 2011 tests administered to
representative samples of U.S. students in 8th grade by the National
Assessment of Educational Progress (NAEP). Our country-by-country data
come from the PISA tests, which are administered by the Organization for
Economic Co-operation and Development (OECD). In 2012, OECD administered
the PISA tests to representative samples of students at the age of 15 in
68 jurisdictions, including all 34 OECD countries. Our analysis compares
U.S. performance to those of students in the other OECD countries.
The proficiency and advanced standards used in this study follow
those developed by NAEP. To equate proficiency and advanced performance
rates across states and countries, we execute a crosswalk between the
NAEP and PISA tests by identifying levels of performance on PISA that
yield equivalent proportions of U.S. students that meet the NAEP
proficiency and advanced standards (see Methodology sidebar).
To assess overall performance, we identify the percentage of
students in the high school class of 2015 who are performing at
proficient and advanced levels of achievement in math. (While not
reported here, we also looked at reading and science, and the results
are broadly similar to those for math.) We focus on how each state
within the United States ranks relative to all 33 other OECD countries.
To ascertain whether the challenges facing the United States are
concentrated among the educationally disadvantaged, we identify for each
state and country the proficiency rate of students from families with
parents of high, moderate, and low levels of education. If the problems
are concentrated in ways that some would have us believe, U.S. students
from families with high parental education should compare favorably with
similarly situated students abroad. Such a finding would support the
oft-repeated claim that the achievement challenges are limited to those
who come from disadvantaged families (measured here by low levels of
parental education).
How Well Do U.S. Schools Educate Different Students?
According to NAEP, 35 percent of the members of the U.S. class of
2015 reach or exceed the proficiency level in math. Based on our
calculations, this percentage places the United States at the 27th rank
among the 34 OECD countries (see Figure 1). The percentage of students
who are math proficient is nearly twice as large in Korea (65%), Japan
(59%), and Switzerland (57%). Other countries with performances that
clearly outrank the United States include Finland (52%), Canada (51%),
Germany (50%), Australia (45%), France (42%), and the United Kingdom
(41%).
Student Proficiency Overall (Figure 1)
Percentage at or above proficiency level in math
among all students in the Class of 2015 in U.S.
states and OECD countries.
Rank Rank Political % proficient
Among Among Jurisdiction in U.S. 1
Countries States
1 Korea 65.0%
2 Japan 59.2
3 Switzerland 57.3
4 Netherlands 54.7
5 Finland 51.8
6 Estonia 51.3
7 1 Massachusetts 51.2
7 Canada 51.0
8 Belgium 50.3
9 Germany 50.0
10 Poland 49.3
11 2 Minnesota 47.6
11 3 New Jersey 46.8
11 Austria 46.3
12 4 Vermont 46.0
12 5 Montana 45.6
12 Australia 45.0
13 Czech Republic 43.7
14 Ireland 43.6
15 6 New Hampshire 43.6
15 7 Colorado 43.5
15 New Zealand 43.4
16 Slovenia 42.8
17 Denmark 42.8
18 8 North Dakota 42.6
18 France 42.4
19 9 South Dakota 41.7
19 United Kingdom 41.3
20 10 Wisconsin 41.0
20 11 Kansas 40.8
20 Iceland 40.7
21 12 Washington 40.4
21 13 Maryland 40.4
21 Luxembourg 40.3
22 14 Texas 40.0
22 15 Virginia 39.7
22 Norway 39.3
23 16 Ohio 38.9
23 17 Pennsylvania 38.9
23 Portugal 38.8
24 18 Maine 38.8
24 19 Connecticut 38.1
24 20 Wyoming 37.4%
24 Italy 37.4
25 Slovak Republic 37.11
26 21 North Carolina 37.0
26 Spain 36.9
27 22 Idaho 36.9
27 23 Alaska 35.2
27 24 Utah 34.9
27 United States 34.7
28 Sweden 34.6
29 25 Indiana 34.1
29 26 Rhode Island 33.9
29 27 Iowa 33.6
29 Israel 33.3
30 Hungary 33.1
31 28 Illinois 32.8
31 29 Nebraska 32.8
31 30 Oregon 32.7
31 31 Delaware 31.9
31 32 South Carolina 31.8
31 33 Missouri 31.5
31 34 Arizona 31.5
31 35 Michigan 30.8
31 36 Kentucky 30.7
31 37 New York 30.0
31 38 Hawaii 30.0
31 39 Arkansas 29.3
31 40 Nevada 28.6
31 41 Georgia 27.8
31 42 Florida 27.7
31 43 Oklahoma 27.3
31 44 California 25.3
31 Greece 24.0
32 45 Tennessee 23.9
32 46 New Mexico 23.8
32 Turkey 23.9
33 47 Louisiana 22.3
33 48 West Virginia 21.3
33 49 Alabama 20.1
33 50 Mississippi 19.3
33 Chile 13.5I
34 Mexico 8.8
NOTE: States ranked against the OECD countries
without displacing any countries in
SOURCE: Author? calculations
the rank order and without regard to the
position of other states.
To see whether the low U.S. ranking in math is due mainly to social
class factors separate and apart from the schools, we next identify
proficiency ratings for students from families with differing amounts of
parental education.
Low parental education. Only 17 percent of these U.S. students are
proficient in math (see Figure 2). This is half' or less than the
percentage of similarly situated students (those whose parents also have
low levels of education) in Korea (46%), the Netherlands (37%), Germany
(35%), and Japan (34%). Among OECD countries as a whole, the United
States ranks 20th, placing it slightly ahead of Austria and France and
just behind Denmark and the United Kingdom. In simplest terms, many
other countries do a much better job of educating young people whose
parents lack a high school diploma.
Student Proficiency, Low Parental Education
(Figure 2)
Percentage at or above proficiency level in math
among students whose parents have a low level of
education in the Class
Rank Rank Political % proficient
Among Among Jurisdiction in U.S.
Countries States
1 Korea 45.7%
2 Netherlands 36.9
3 Germany 34.6
4 Japan 33.6
5 Switzerland 32.6
6 Finland 27.7
7 1 Texas 27.6
7 Estonia 27.4
8 Portugal 26.6
9 Australia 25.8
10 Iceland 25.2
11 Belgium 25.0
12 2 New Jersey 24.8
12 Ireland 23.2
13 Canada 23.1
14 3 New Hampshire 22.4
14 Italy 22.1
15 4 North Carolina 22.1
15 Spain 20.8
16 5 Montana 19.7
16 New Zealand 19.0
17 Luxembourg 18.8
18 United Kingdom 18.8
19 6 Kansas 18.6
19 7 Massachusetts 18.5
19 8 Minnesota 17.9
19 Denmark 17.5
20 9 Delaware 17.4
20 10 Washington 17.0
20 United States 16.7
21 11 Maryland 16.7
21 12 Arizona 16.6
21 13 Maine 16.4
21 Austria 16.3
22 France 16.3
23 14 Illinois 16.3
23 15 Arkansas 16.2
23 16 Kentucky 16.0
23 17 Indiana 16.0
23 Sweden 15.7
24 Poland 15.7
25 18 Wyoming 15.6
25 19 Hawaii 15.5%
25 20 South Dakota 15.1
25 21 Virginia 14.9
25 Turkey 14.4
26 Norway 14.0
27 22 Vermont 13.9
27 23 Florida 13.5
27 24 Idaho 13.1
27 25 Georgia 12.9
27 Slovenia 12.4
28 26 Rhode Island 12.4
28 27 New York 12.2
28 28 Oregon 12.1
28 29 Nevada 12.0
28 30 Ohio 11.9
28 Czech Republic 11.9
29 31 Connecticut 11.6
29 32 Nebraska 11.3
29 33 Colorado 11.3
29 34 Louisiana 11.3
29 35 South Carolina 10.6
29 36 Tennessee 10.6
29 37 Missouri 10.5
29 38 Mississippi 10.4
29 39 Pennsylvania 10.2
29 40 Iowa 10.1
29 41 Wisconsin 9.9
29 42 Oklahoma 9.8
29 43 Michigan 9.5
29 44 Alabama 9.2
29 45 New Mexico 9.1
29 46 California 8.8
29 Greece 8.6
30 Hungary 7.2
31 Israel 6.3
32 47 West Virginia 6.2
32 48 Utah 5.4
32 Mexico 4.7
33 Slovak Republic 2.3
34 Chile 2.3
NOTES: See note in Figure
SOURCE; Authors calculations
1. No data are available for Alaska and North
Dakota.
Moderate parental education. The relative standing of the United
States is even worse among students from moderately well educated
families. The math proficiency rate (26%) for this group is again around
half the rate enjoyed by Switzerland (57%), Korea (56%), Germany (52%),
and the Netherlands (50%). Other major countries that outperform the
United States include Japan (48%), Canada (43%), Poland (43%), the
United Kingdom (39%), and France (35%). When it comes to instructing the
children of the moderately well educated, the United States comes in at
the 30th rank among the 34 OECD countries, 10 ranks lower than was the
case for students from families with low parental education.
High parental education. The percentage proficient of 15-year-olds
from families with high parental education is conventionally thought to
be the exception to this bleak picture. Indeed, the proficiency rate of
43% is higher than the rate for families with low (17%) or moderate
(26%) levels of education. But the relative standing of the United
States vis-a-vis other OECD countries remains near the very bottom (see
Figure 3), at the 28th rank. When viewed from a global perspective, U.S.
schools seem to do as badly teaching those from better-educated families
as they do teaching those from the less well educated.
Student Proficiency, High Parental Education (Figure 3)
Percentage at or above proficiency level in math
among students whose parents have a high level of
education in the Class of 2015 in US. states and
OED countries.
Rank Rank Political % proficient
Among Among Jurisdiction in U.S.
Countries States
1 Korea 72.8%
2 Poland 71.2
3 Japan 67.9
4 Switzerland 64.9
5 Germany 63.7
6 1 Massachusetts 62.3
6 Netherlands 60.9
7 Belgium 60.6
8 2 Vermont 59.3
8 Portugal 59.1
9 3 Minnesota 59.0
9 Czech Republic 38.2
10 4 Colorado 58.1
10 5 New Jersey 57.9
10 6 Montana 57.5
10 Estonia 56.6
11 Slovenia 56.5
12 Canada 56.6
13 Finland 55.8
14 New Zealand 55.3
15 France 55.3
16 Australia 55.2
17 Austria 54.8
18 7 Washington 54.3
18 8 Texas 54.2
18 Slovak Republic 54.0
19 Luxembourg 53.1
20 9 New Hampshire 53.0
20 Ireland 52.9
21 10 Virginia 52.6
21 11 Wisconsin 52.6
21 12 Kansas 52.2
21 13 Maryland 52.0
21 14 South Dakota 51.6
21 15 Connecticut 51.3
21 16 Pennsylvania 50.7
21 17 North Dakota 50.2
21 Denmark 50.0
22 18 Ohio 49.8
22 19 Idaho 49.6
22 20 Maine 49.4
22 21 Arizona 49.1
22 United Kingdom 48.5%
23 22 Wyoming 48.2
23 23 North Carolina 48.1
23 24 Rhode Island 48.0
23 25 Utah 47.8
23 Spain 47.1
24 26 Indiana 46.9
24 27 Oregon 46.9
24 Hungary 46.6
25 Iceland 45.7
26 28 Illinois 45.6
26 Norway 45.1
27 Israel 45.0
28 29 Iowa 44.9
28 30 Nebraska 44.8
28 31 Kentucky 44.1
28 United States 18.3
29 32 South Carolina 43.1
29 Italy 42.6
30 33 California 42.6
30 34 Missouri 42.3
30 35 Michigan 42.0
30 36 Oklahoma 41.8
30 37 Nevada 41.7
30 38 Delaware 40.9
30 Turkey 40.5
31 39 Arkansas 40.2
31 40 New York 39.8
31 Sweden 39.1
32 41 Georgia 38.2
32 42 Hawaii 37.5
32 43 Florida 37.5
32 44 New Mexico 37.1
32 45 Tennessee 34.1
32 Greece 32.3
33 46 West Virginia 31.9
33 47 Louisiana 28.1
33 48 Alabama 27.8
33 Chile 26.0
34 49 Mississippi 25.6
34 Mexico 14.1
NOTES: See note In
SOURCE: Authors'caku!ation5
Figure 1. No data are available for Alaska.
Countries with high proficiency rates among students from
better-educated families include Korea (73%), Poland (71%), Japan (68%),
Switzerland (65%), Germany (64%) and Canada (57%). Perhaps the only
comfort the United States can take is that it is only 5 percentage
points behind its mother country, the United Kingdom (48%).
Across the OECD, there is a strong relationship between the math
performance of students from families with high and with low educational
backgrounds. Mexico and Chile are particularly weak at educating those
from better-educated families, however. Conversely, Poland and Slovakia
are particularly weak at educating students from families with less
education, given the performance of those from families with high
education. The relative performance of the U.S. education system is
pretty much the same across social groups. It is weak at the bottom, no
less weak at the middle, and just as weak with respect to educating the
most-advantaged. As Secretary Duncan said, it is not a problem of some
other person's child.
Ranking States
The overall math proficiency rate of 15-year-olds varies widely
among the states--from a high of 51 percent in Massachusetts to a low of
19 percent in Mississippi. Striking differences remain when one divides
students according to parental education. For students from families
with low parental education levels, Texas (28%) and New jersey (25%)
have the highest proficiency rates, well ahead of Massachusetts and
Minnesota (both at 18%), putting them in 7th and 8th place among U.S.
states for this category of students. Maryland and Illinois are at about
the national average, while New York, in 27th place, falls slightly
below. California (9%), West Virginia (6%), and Utah (5%) rank at
embarrassingly low levels. (See Figure 4 for a picture of the overall
pattern throughout the 50 states.)
Many people assume that students coming from families with high
education levels are keeping up with their peers abroad. Indeed, in some
parts of the United States that is in fact the case. More than 62
percent of students from Massachusetts families with high levels of
parental education are proficient in math, placing that state just
behind Germany (64%) and Switzerland (65%), two of the top-five OECD
countries. Only a bit further back are Vermont, Minnesota, Colorado, New
Jersey, and Montana, all of which have a proficiency rate of 58 percent
or 59 percent for students from better-educated families.
Internationally, that places these states in the same league as the
Czech Republic (58%), Canada (57%), and Finland (56%), which are among
the OECD top 13.
But those six states are the highest-performing states in the
Union. Other states rank much lower down the international list. In many
places, students from highly educated families are performing well below
the OECD average for similarly advantaged students. For example,
Wisconsin, if ranked as a country on this measure, would come in 21st,
just below Ireland. California is large enough to be an OECD country in
its own right. If it were, its 43 percent proficiency rating would place
it 30th, just below Italy, and New York's 40 percent rating
entitles it to assume position number 31, just below Turkey.
Florida's 38 percent rating gives it the 32nd position, just below
Sweden, which has registered an abysmal performance given its level of
economic development. Ranked near the bottom, Alabama, West Virginia,
and Louisiana do worse than all OECD countries with the exception of
Chile and Mexico. (See Figure 5 for an overall portrait of the pattern
among the states.)
Similar to the international comparisons, states that rank well for
math education among students with high parental education tend also to
rank highly for students from less-advantaged backgrounds. But some
high-performing states, such as Massachusetts, Vermont, and Colorado, do
relatively better with students from families with higher educational
backgrounds than they do with their less-advantaged peers.
Advanced Performance in Math
The U.S. economic strength has been built in large part through its
record of invention and innovation, things that themselves depend upon
the country's historic strength in science, technical, engineering,
and math fields (STEM). The pool of people prepared to go into these
fields in the future is dependent on students who have developed
advanced skills in math and science in school.
[ILLUSTRATION OMITTED]
Eight percent of the U.S. class of 2015 proved its merit by scoring
at the advanced level on the NAEP in math. That could be regarded as a
triumph were it not for the fact that it leaves the United States 28th
on the OECD list. Other countries do a much better job at bringing
students up to the advanced level of performance. The eight world
leaders are Korea (30%), Japan (23%), Switzerland (20%), Belgium (19%),
the Netherlands (18%), Germany (17%), Poland (16%), and Canada (16%).
Disturbingly, our neighbor to the north turns out twice as high a
percentage of students at the advanced level in math as the United
States.
The percentage scoring at the advanced level is only 2 percent for
U.S. students from families with low levels of educational attainment
and only 4 percent for students from moderately educated families. Those
disgraceful numbers could be offset by unusually high performances among
the better-educated, however. Does the United States achieve a
breakthrough at least among this group? Some may wish to take pride in
the fact that 12 percent of the students from better-educated families
reach the advanced level in math. But such pride is misplaced, as the
feat still leaves the United States in the 28th position out of the 34
OECD countries. Only Sweden, Spain, Turkey, Greece, Chile, and Mexico do
worse.
[ILLUSTRATION OMITTED]
[ILLUSTRATION OMITTED]
Advanced Performance by State
The four states with 13 percent or more students performing at the
advanced level in math are Massachusetts, New Jersey, Minnesota, and
Vermont, with the Bay State taking honors with 15 percent of its
students scoring at that level. All of these states rank alongside the
top 13 OECD countries, and Massachusetts ranks 9th, just below Canada,
though still well below Korea and Japan. With less than 7 percent of
students performing at the advanced level, New York and California rank
31st, just ahead of Turkey and Greece. The two lowest-performing states,
Alabama and Louisiana, however, do outrank the two lowest-performing
OECD countries--Chile and Mexico.
The same states--Massachusetts, New Jersey, Minnesota, and
Vermont--are top performers on this measure for students from families
with high educational backgrounds; in all four plus Colorado, 18 percent
or more of such students perform at the advanced level. That places them
in the same league as Canada and France but well behind Korea, Poland,
Japan, Switzerland, Belgium, and Germany. But only 15 percent perform at
this level in Pennsylvania and 14 percent in Wisconsin, and less than 10
percent do so in New York, Michigan, and Florida. If states do well with
students from better-educated family backgrounds, they tend to do well
with those from less-educated ones. But there are clear exceptions to
this pattern. West Virginia, Louisiana, and Mississippi score
particularly badly on their capacity to teach students from
more-educated backgrounds.
Conclusions
Lacking good information, it has been easy even for sophisticated
Americans to be seduced by apologists who would have the public believe
the problems are simply those of poor kids in central city schools. Our
results point in quite the opposite direction. We find that the
international rankings of the United States and the individual states
are not much different for students from advantaged backgrounds than for
those from disadvantaged ones. Although a higher proportion of U.S.
students from better-educated families are proficient, that is equally
true for similarly situated students in other countries. Compared to
their counterparts abroad, however, U.S. students from advantaged homes
lag severely behind.
As long as the focus remains on distinctions within the United
States, then the comfortable can remain comforted by the distance
between suburbia and the inner city. But once the focus shifts to
countries abroad and fair, apples-to-apples comparisons are made, it
becomes manifest that nearly all of our young people--from privileged
and not-so-privileged backgrounds--are not faring well.
Some say that we must cure poverty before we can address the
achievement problems in our schools. Others say that our schools are
generally doing fine, except for the schools serving the poor. Bringing
an international perspective correctly to bear on the issue dispels both
thoughts.
The United States has two achievement gaps to be bridged--the one
between the advantaged and the disadvantaged and the one between itself
and its peers abroad. Neither goal need be sacrificed to attain the
other.
RELATED ARTICLE: Methodology
Our analysis relies on test-score information for adolescents
collected by NAEP in 2011 and PISA in 2012. To equate proficiency and
advanced performance rates across states and countries, we execute a
crosswalk between the two tests by identifying levels of performance on
PISA that yield equivalent proportions of U.S. students that meet the
NAEP proficiency and advanced standards. We assume that all those who
pass the NAEP proficiency bar in 8th grade will pass a similar threshold
on the PISA test the next year. The 2011 NAEP assessment identifies 34.7
percent of U.S. 8th graders as proficient and 8.2 percent as advanced in
math. Thus, in math, that threshold is calculated by identifying the
lowest PISA score of students who rank in the top 34.7 percent of U.S.
PISA test-takers. Similar procedures are used to conduct crosswalks at
the advanced level in math.
Low levels of parental education are defined here as having no
parent who received a high school diploma. Families with moderate
education levels are those in which at least one parent is reported to
have received a high school diploma but neither parent has earned a
college degree. Families with high education levels are those reported
to have at least one parent with a college degree. (See the full report
for further methodological details. Available at
http://www.hks.harvard.edu/pepg/PDF/Papers/PEPG14_01_NotJust.pdf.)
ERIC A. HANUSHEK, PAUL E. PETERSON, and LUDGER WOESSMANN
Eric A. Hanushek is senior fellow at the Hoover Institution of
Stanford University. Paul E. Peterson is professor of government and
director of the Program on Education Policy and Governance at Harvard
University. Ludger Woessmann is professor of economics at the University
of Munich and director of the Ifo Center for the Economics of Education
and Innovation. An unabridged version of this report is available at
hks.harvard.edu/pepg/.
