Internet Subsidies in Public Schools Research by Austan D. Goolsbee and Jonathan Guryan
As Internet use has dramatically increased over the past five
years, policymakers have grown concerned about the so-called
"digital divide," in which access and understanding
of technology would be split across income and racial lines.
New research analyzes the impact of an ambitious federal technology
program aimed to bridge the digital divide in public schools.
Policymakers and analysts have argued that public schools are
a natural place to teach underserved populations about computers
and increase access to new technology. As part of the Telecommunications
Act of 1996, the government began actively subsidizing Internet
and telecommunications access in U.S. classrooms and libraries
through a tax on long-distance telephone services.
Starting in 1998 and continuing today, the E-Rate program provides
$2.25 billion annually in subsidies to schools and libraries
so they can invest in Internet and communications technology.
This is especially significant when compared to the combined
computer budget in all public schools, which totaled only $3.3
billion in 1999.
In their new study, "The Impact of Internet Subsidies
in Public Schools," University of Chicago Graduate School
of Business professors Austan D. Goolsbee and Jonathan Guryan
provide the first step towards evaluating the effect of the
E-Rate subsidy on Internet investment in California public schools.
"The effect of differing access to technology can magnify
as people get older, so the subsidy may function as preventive
medicine," says Guryan.
Goolsbee and Guryan approached their study with two goals in
mind: 1) to understand the extent to which the $2.25 billion
subsidy fulfilled its primary goal of increasing Internet usage
in schools, particularly those serving disadvantaged students,
versus merely subsidizing spending that was already taking place;
and 2) addressing the considerable interest in and debate about
whether spending on computers and information technology has
any impact on student performance.
Employing new data on school technology usage in every California
public school from 1996 to 2000, as well as application data
from the E-Rate program, Goolsbee and Guryan find that the subsidy
did succeed in significantly increasing schools' investment
in Internet technology. By 2000, there were approximately 66
percent more classrooms in California with Internet connections
than there would have been without the subsidy-the equivalent
of accelerating Internet investment by about four years.
Despite the noticeable increase in classroom Internet connections,
the authors find very little evidence that the program has any
impact on student achievement, as measured by test scores in
a variety of subjects.
Analyzing the E-Rate
The E-Rate program was designed to give all eligible schools
and libraries affordable access to modern telecommunications
and information services. The subsidy can be used for spending
on all commercially available telecommunications services, Internet
access, and internal communications, but not for buying computers
for the school. The subsidy rate ranges from 20 to 90 percent
depending on the school's share of students that qualify for
the national school lunch program.
The federal government subsidizes or gives free school lunches
to students whose family income is below a certain level, usually
close to the poverty level. The number of school lunch eligible
students determines funding for many federal and state education-related
programs. The higher that number is, the higher the subsidy
By 1997, the year before the first E-Rate funding was awarded,
55 percent of California's public schools had at least one classroom
with Internet access, an increase of 9 percentage points from
1996. Even without subsidies, many school districts chose to
make Internet investments. It is therefore difficult to distinguish
between the effects of the E-Rate program alone versus the already
strong upward trend in the fraction of schools with Internet
To address this concern, Goolsbee and Guryan used new data
on the information technology owned in each year by every school
in California. They merged this with administrative data on
every E-Rate funding application these same schools filed, including
the amount of funding, the subsidy rate, and the purpose of
the funding. They also included demographic data for each school
and district from the National Center for Education Statistics
Common Core of Data and the 1990 U.S. Census.
California was chosen primarily because the state was already
measuring technology usage at schools before the E-Rate program
"The fact that we were able to get detailed data on individual
schools and what they had purchased before the E-Rate began
is what allowed us to identify the impact of the program separately
from just underlying trends in the economy," says Goolsbee.
From 1996 to 1997, the richest California schools had almost
50 percent more Internet classrooms per teacher than the poorest
group. Over the next year, this gap increased. Once the E-Rate
program began in 1998, however, the poorer groups began receiving
large subsidies relative to the richest group, making Internet
connections much more affordable. Their relative number of Internet
connections accelerated and by 2000, some of the poorest school
districts actually had more Internet connections than the richest
Goolsbee and Guryan find that most of the E-Rate funds went
to schools with subsidy rates of 80 to 90 percent-schools with
more than 50 percent of their students eligible for free or
reduced price lunches from the federal government.
This reversal in the Internet access rates of rich and poor
schools indicates a closing of the digital divide that occurred
in the years following the introduction of the E-Rate program.
Rural schools are substantially less responsive to the subsidy
than urban schools. The subsidies had the greatest impact on
schools with heavily black and Hispanic student populations.
Schools with a majority of white and Asian students did not
show a significant increase in Internet connections. Elementary
schools also respond to the subsidy to a greater degree than
"As a technology subsidy, the E-Rate program is a tremendous
success," says Goolsbee. "Today, the majority of classrooms
in schools are connected to the Internet."
Gauging Student Performance
Having established that the E-Rate program achieved its primary
goal of getting classrooms connected to the Internet, particularly
at disadvantaged schools, Goolsbee and Guryan extended their
evaluation to the next logical level: Does this investment in
technology have any impact on student performance?
Many of the most prominent supporters of the E-Rate program
argued that the program should aspire to more than just wiring
schools, and should help students develop basic skills.
Goolsbee and Guryan measured student achievement using the
Stanford Achievement Test, which has been given every year to
all elementary, middle, and high school students in California
public schools beginning with the 1997-98 school year. They
used three measures of school-level achievement: the average
test score in the school, the fraction of students scoring above
the 75th percentile score for the state, and the fraction of
students scoring above the 25th percentile score for the state.
The authors looked at test scores for math, reading, and science.
Their results showed no evidence that investments in Internet
technology had any measurable effect on student achievement.
However, they note that it may be too early to evaluate long-term
investments in information technology, or that gains may have
taken place in areas other than test scores.
The lack of impact on student performance is consistent with
Department of Education evidence that only one-third of teachers
reported they were prepared or well-prepared to use computers
and the Internet in the classroom. If teachers are uncomfortable
working with the technology or computers in general, it is easy
to see why greater Internet access may not directly result in
better prepared students.
"It certainly would have been great if you spent a modest
amount of money on this program and saw a big educational bang
for the buck," says Goolsbee. "But for researchers
it's not surprising that the subsidy did not have this effect,
because there is no magic bullet for improving student performance."
Making the Internet Work in Classrooms
Goolsbee and Guryan suggest the next step for studying effects
of the E-Rate program would be to figure out why the high rates
of Internet adoption have had a minimal impact on student performance.
"One possibility would be to run pilot programs, collect
data, and see how schools are using the technology," says
Guryan. "Are there schools that introduce the technology
very effectively? What kind of schools are they? What are the
skills that teachers need to be able to use the Internet?"
The authors highlight the importance of finding out what the
technology actually achieves in the classroom, and whether teachers
are even incorporating the Internet into the curriculum. Getting
classrooms connected is only the first hurdle toward eliminating
the digital divide.
Austan D. Goolsbee is professor of economics at the University
of Chicago Graduate School of Business. Jonathan Guryan is assistant
professor of economics at the University of Chicago Graduate
School of Business.