Formula for computing chi square: |
 |
| S means summation - add them up. |
SAN DIEGO (AP) -- San Diego
police officers are
more likely to stop and
search blacks and Hispanics
than whites and Asians,
according to a study for the
police
department.
Researchers found that whites
and Asians had an 8
percent chance of being
pulled over in traffic stops
during the first six months
of this year. Blacks and
Hispanics had a 14 percent
chance of being stopped,
they found.
``This is no surprise to
us. We live this reality every
day,'' said Sydney Etheridge,
president of San Diego's
Coalition of African-American
Organizations.
Police Chief David Bejarano
promised further
investigation but said Thursday
that he's seen nothing to
make him believe his 2,100
officers are acting
improperly.
Bejarano, the city's first
Latino chief, ordered the
$100,000 study last year
after complaints that officers
stop blacks and Hispanics
for minor infractions that
would be overlooked if committed
by a white or Asian.
The study was based on data
collected by officers
during 91,552 traffic stops
in the first six months of the
year. Researchers at three
universities compared the
ethnicity of the people
stopped with the numbers of
driving-age people in each
ethnic group.
Allegations of racial profiling
by police have become a
volatile political issue
around the country. Five states
have outlawed the tactics
and dozens of police
departments are gathering
ethnic statistics on people
who are stopped and searched.
California Gov. Gray Davis
last year vetoed a bill that
would have required law
enforcement agencies to keep
ethnic data on traffic stops,
saying decisions about
gathering such statistics
are best left to local agencies.
NYT Oct 1, 2000
San
Diego Police Found to Stop Black and Latino Drivers Most
By BARBARA WHITAKER
LOS ANGELES,
Sept. 30 — African-Americans and
Hispanics
have a much greater chance of being
stopped by the police in
San Diego than whites and
Asian-Americans, according
to the results of a survey
released this week.
The survey, conducted for
the San Diego Police
Department in the first
six months of the year, found that
whites and Asian-Americans
had an 8 percent chance
of being pulled over while
African-Americans and
Hispanics had a 14 percent
chance. The chances were
calculated by comparing
the ethnic breakdown of those
stopped in traffic with
the number of people of driving
age in San Diego.
The survey found that while
blacks are 8 percent of the
driving age population in
San Diego, they accounted for
12 percent of the traffic
stops and 20 percent of those
searched. Hispanics, who
are 20 percent of the driving-
age population, made up
29 percent of the stops and 50
percent of the searches.
"We have these numbers, and
the numbers are
troubling," said Dale Kelly
Bankhead, a spokeswoman
for the American Civil Liberties
Union of San Diego
and Imperial Counties.
But Ms. Bankhead applauded
the department for
studying the issue, which
has been raised in cities
across the country, calling
the department the first to
study racial profiling voluntarily.
At a news conference on Thursday
to announce the
results, Police Chief David
Bejarano said there was
nothing in the initial findings
that made him believe
officers were acting improperly.
And in a videotape
shown to the department's
2,100 officers, Mr. Bejarano
said they should not take
the results as a negative
reflection on their work.
However, Mr. Bejarano, the
city's first Hispanic police
chief, said the numbers
did indicate that more study
was needed. The numbers
announced this week were
preliminary figures, and
the study will continue for the
rest of the year.
The authorities said the
results could have been skewed
by several factors. For
example, analysts used the 1990
Census to determine the
number of people of driving
age in San Diego, but critics
contend that minorities are
underreported in the survey.
Also because San Diego
borders Mexico, many Mexicans
drive over every day,
but would not be counted
in the Census numbers.
Still, Mr. Bejarano said
that the study had raised
serious questions and that
if abuses were found they
would be addressed. "If
there are problems," he said,
"we want to make sure we
address them with training,
supervision and, if necessary,
discipline."
Racial profiling, a practice
where the police stop
motorists based on their
race, has become a contentious
issue nationwide. At least
five states have outlawed the
practice and at the suggestion
of United States Attorney
General Janet Reno, the
police are beginning to gather
statistics on stops and
searches. Still, many
departments, including the
Los Angeles Police
Department, balk at the
suggestion.
The San Diego police began
keeping records of stops
and searches in January
after concerns were raised by
community leaders. For more
than 91,000 stops, the
police recorded the date
and time, reason, driver's sex,
age and race and action
taken.
The information was then
analyzed by academics from
Eastern Kentucky University,
Vanderbilt University and
San Diego State University
with the police and
representatives from groups
like the N.A.A.C.P., Urban
League and A.C.L.U. reviewing
the results.
Question: How many stops should there be, on the assumption that people of all groups are equally likely to commit traffic offenses?
The U.S. Census's State and Metropolitan Area Data Book gives the following figures for the San Diego, CA Metropolitan Statistical Area (p 70 of the text, 82 in Adobe Acrobat). These might not be the best baseline statistics to use, you might want city figures or you might want licensed drivers - but lots of people from the suburbs drive through the city. For our example, let's use these statistics:
Population estimate for 2000: 2,963,900
Percent (1996) of population White 82.5%, black 6.4%, Asian or
pacific islander 10.2%, Hispanic Origin (may be of any race) 25.1%.
Given these statistics, how many white people lived in San Diego MSA?
How many black people?
How many Hispanic people?
Now, on the assumption that blacks, white and hispanics should be receiving
the same proportion of traffic stops, how many of 91,000 traffic stops
should be for each group? Put those numbers in the "Expected" column.
Also calculate the "observed percent" based on the Observed Frequencies
(which I made up for this example).
| Population Groups | Expected | Observed frequencies | Observed Percent |
| White (82.5% of pop.) | 71,718 | ||
| Black (6.4%) | 11,182 | ||
| Asian (10.2%) | 7,495 | ||
| Other (0.9%) | 605 | ||
| Total | 91,000 | 91,000 | 100% |
Now, are those "expected" numbers significantly different from the Observed Numbers? To find out, solve the formula for Chi Square [in any statistics book or in the box below, detailed computing instructions are also available online] and look the result up in the table of values of the chi square distribution in the back of a statistics book..(or online).. OR... use a handy WEB chi square calculator at Georgetown University (or, for expected and observed problems only, not cross-tabulations, the one at Vassar; there are others also).
What is the chi square value?
| Formula for computing chi square: |
|
| S means summation - add them up. |
Is this finding statistically significant?
Would this be considered evidence for or against allegations of racial profiling?
-----------------------
NYT Oct 7, 2000
A similar case in Texas:
Studies Find Race Disparities in
Texas Traffic Stops
By JIM YARDLEY
HOUSTON,
Oct. 6 — Black and
Hispanic motorists across
Texas are more than twice
as likely
as non-Hispanic whites to
be
searched during traffic
stops while
black drivers in certain
rural areas
of the state are also far
more likely
to be ticketed, according
to two
studies examining possible
racial
profiling.
The investigations by the
National Association for the
Advancement of Colored People
and by The Dallas
Morning News were released
this week and brought an
immediate challenge from
the Texas Department of
Public Safety. State officials
described the Morning
News study as flawed, citing
other statistics in denying
that the agency practiced
racial profiling.
The investigations were undertaken
after complaints in
Texas and other states.
Gary Bledsoe, state president
of the N.A.A.C.P., said
he believed that the two
reports proved that minority
drivers were singled out
by many law enforcement
officers. "The data are
very clear," said Mr. Bledsoe,
an Austin lawyer who said
he had been stopped but not
ticketed nearly 20 times
during the past decade. "This
is a clear indication of
racial profiling. It's definitely
happening."
The two studies used the
same raw data but focused on
different time periods and
used different experts to
tabulate the numbers. The
studies centered on two
fundamental issues: the
number of minorities ticketed
and the number of minorities
searched during traffic
stops.
The Morning News, which published
its investigation
on Wednesday, examined roughly
895,000 traffic
tickets written by state
troopers last year. The figures
were analyzed by a University
of Texas mathematics
professor.
Looking at the state as a
whole, the newspaper
concluded that blacks and
Hispanics received tickets at
rates that actually were
proportional to their
driving-age populations.
But disparities emerged when
the study examined different
regions of the state. In
many rural counties blacks
were nearly twice as likely
as non-Hispanic whites to
be ticketed. Of the 193
counties analyzed, blacks
received more tickets than
expected in 84 counties.
There were some counties
where whites received a
higher- than-expected number
of tickets, but in those
cases the disparity was not as
great.
James B. Francis, board chairman
of the Department of
Public Safety, said the
Morning News analysis was
flawed because it compared
the race and number of
ticketed drivers with the
local population where the
stop occurred. But, he said,
it did not consider that the
drivers might be from elsewhere.
"I'm not going to start
a massive investigation
unless and until there is some
indication that something
is going on," Mr. Francis
said.
On Tuesday, Mr. Bledsoe released
the N.A.A.C.P.
findings at a news conference
in Austin. The report,
prepared by two professors
of economics and statistics,
analyzed 65,000 traffic
stops from March and found
that blacks were searched
twice as often as
non-Hispanic whites while
Hispanics were searched
two and a half times as
often. The study also concluded
that search rates were even
higher for minority males
— with Hispanic men four
times as likely as
non-Hispanic whites to be
searched and black men two
and a half times as likely.
Mike Cox, a public safety
spokesman, declined to
comment and steered any
questions to a report about
traffic stops posted this
week on the agency's Web site.
In March, the department
began collecting data on
traffic stops, requesting
that officers collect information
like a driver's race and
sex. The initiative began after
the public safety department
suspended seven officers
in East Texas for racial
insensitivity. After this
incident, Mr. Francis, the
board chairman, declared
there would be "zero tolerance"
for discrimination
within the ranks.
The statistics on the internal
report, which are based on
five months of data from
this year, reveal two primary
findings, one suggesting
that racial profiling does not
occur and the other suggesting
that it does. First, the
state report found that
non-Hispanic whites constituted
a higher rate of overall
traffic stops (68.12 percent)
than their estimated statewide
population (60.69
percent). Blacks and Hispanics,
the report states, are
actually stopped at lower
rates than their overall
populations.
But the state statistics
also show that blacks and
Hispanics are twice as likely
as non-Hispanic whites to
be searched during stops,
a finding similar to those of
the N.A.A.C.P. and The Morning
News. The state
report explained the high
number of Hispanics who
were searched as a byproduct
of the traffic of illegal
immigrants and drugs from
Mexico.