By Natasha Singer, The New York Times
The number of trackers collecting data on users’ activities on the
most popular Web sites in the United States has significantly increased
in the last five months, according to research from the Berkeley Center
for Law and Technology at the University of California, Berkeley.
The Berkeley project, called the “Web Privacy Census,” aims to
measure online privacy by conducting periodic web crawls and comparing
the number of cookies and other types of tracking technology found over
time on the most visited sites.
During a test conducted on Oct. 24, researchers encountered cookies
on every site included in a list of the 100 most popular sites compiled
by Quantcast, an analytics and audience targeting firm.
On those 100 sites, researchers found 6,485 standard cookies last
month compared with 5,795 cookies in May. In both months, third party
trackers, not the Web sites themselves, set a majority of those cookies,
the report said.
In October and May, cookies placed by DoubleClick, Google’s ad
technology service, appeared on the most sites on the top 100 list.
ScorecardResearch, an analytics unit of comScore, was the
second-most-prevalent tracker, the researchers reported.
The number of cookies on the top 1,000 and 25,000 Web sites also increased significantly, researchers said.
“More popular sites are using more cookies,” the report said.
The Berkeley study comes at a time of fierce debate among federal
regulators, advertising associations and consumer advocates over how
best to regulate online tracking. Marketers advocate self-regulation,
allowing consumers who wish to opt out of receiving ads based on
data-mining to use an already-established industry program. Some
consumer advocates are pushing for federal regulation as well as a “Do
Not Track” mechanism that would allow Internet users to control tracking
through settings on their own computer browsers.
Chris Hoofnagle, the director of information privacy programs at the
Berkeley center and co-author of the study, said he hoped the data would
set a baseline, providing all sides in the debate with empirical
information as to the optimum method to regulate tracking.
“I’m hoping that it will inform which approach is the best,” Mr.
Hoofnagle said. “We are not going to be well-served unless we measure
these trends more rigorously.”