The Philadelphia
Story
Intro
More than half of
the death sentences rendered in Pennsylvania are cases from Philadelphia,
a city with only 14% of the state's population. Philadelphia's District
Attorney, Lynne Abraham, has been called "The Deadliest D.A."
in a 1995 New York Times article. Eighty-three percent of those on death
row from Philadelphia are African-American. But raw numbers of racial
disproportion do not tell the whole story. In order to determine for
certain whether race is a decisive factor, researchers must examine
the outcomes in cases of similar severity with defendants of similar
criminal backgrounds.
This examination
requires a statistical analysis which takes into account such factors
as multiple victims, the deliberate infliction of pain, and the background
of the accused. The ultimate question is: "Among similar cases,
is race a factor in whether death sentences are imposed against black
defendants?"
Such a study was
recently conducted in Philadelphia. The results are dramatic, particularly
for a state outside of the deep south, a region where racial disparities
in the criminal justice system have a long history. The researchers
examined a large sample of the murders which were eligible for the death
penalty in the state between 1983 and 1993. The researchers found that,
even after controlling for case differences, blacks in Philadelphia
were substantially more likely to get the death penalty than other defendants
who committed similar murders. Black defendants faced odds of receiving
a death sentence that were 3.9 times higher than other similarly situated
defendants.
The results of this
bias against black defendants in Philadelphia is estimated to be an
excess of 38% in death sentences for black defendants compared to all
other defendants for similar crimes.
The
Raw Data
The first step in
determining the presence of racial discrimination in the death penalty
is to look at the raw data: from among the eligible homicides, how often
are black defendants sentenced to death and how often are others sentenced
to death?
The raw data of
death sentences in Philadelphia between 1983 and 1993, provide the first
piece of disturbing evidence that race discrimination may be operating.
The rate at which eligible black defendants were sentenced to death
was nearly 40% higher than the rate for other eligible defendants. In
the chart below, a death sentencing rate of .18 for blacks means that
for every 100 eligible black defendants, 18 will be sentenced to death.
For other defendants, only 13 out of 100 will be similarly sentenced.
Racial disparities
can result through prosecutorial selection of which cases "deserve"
the death penalty, or from the action of juries in determining the final
sentences, or from both. But before a disparity due to race can be established,
a researcher must measure the race effects for crimes of similar severity
committed by defendants with similar criminal histories.
Taking
Into Account the Severity of Murders
Murder cases become
death eligible through the existence of certain aggravating factors
which make one murder "worse" than another. In deciding whether
the death penalty should be sought, the prosecutor is supposed to consider
the presence of such factors as whether a murder was committed with
grave risk to the life of others, whether the murder was committed in
the course of another serious crime such as robbery or rape, whether
torture was used in the commission of the murder, or whether the defendant
had a significant violent history. The jury is similarly told to consider
such factors when deciding whether the sentence should be life or death,
once a guilty verdict is rendered.
Through an analysis
of murders in which the death penalty could have been sought, it is
possible, through an analysis of the defendants that were and were not
sentenced to death, to assign a predictive score, or coefficient, to
various aggravating factors to measure how heavily each influences the
likelihood of a death sentence. The researchers screened hundreds of
factors, statutory and non-statutory, to develop models to explain how
the system works. All statutory factors, and those non-statutory factors
which significantly correlated with the outcome were included.
Comparing the coefficients
permits an average assessment of how much reliance was placed on the
factor by the decision-makers. For example, the fact that the murder
was committed in the course of another felony has less impact than the
fact that the defendant caused great harm, fear or pain. Statistically,
in this study committing another felony had a relative predictive value
of 0.8. On the other hand, if the murder was accompanied by torture,
that factor was very significant and registered a predictive value of
1.9. A murder committed with grave risk of death to others had a relatively
high predictive value of 1.5. A factor which had no apparent effect
would have a value of 0. The study looked at a large class of such variables.
The race of the
defendant is not supposed to influence whether a person is sentenced
to death, but in Philadelphia it clearly does. (See Chart above.) Murders
by blacks are treated as more severe and "deserving" of the
death penalty because of the defendant's race. Being a black defendant
merits a score of 1.4 in predicting whether a death sentence will ultimately
result. This extra burden for black defendants is comparable to such
legitimate aggravating factors as torture or "causing great harm,
fear or pain," which earned scores of 1.9 and 1.0 respectively,
in predicting the sentence. Stated differently, in Philadelphia, the
capital sentencing statute has operated as though being black was not
merely a physical attribute, but as if it were one of the most important
aggravating factors actually justifying the death penalty.
The race of the
defendant is a much stronger predictor that a case will result in a
death sentence than the fact that the crime was committed along with
another felony (0.8) or that the defendant killed with multiple stab
wounds (0.9). Either when the prosecutor decides to seek the death penalty
in a particular case, or when the jury decides that death is the appropriate
sentence, on average, black defendants are considered "worse,"
regardless of the other factors in their case.
Mid-Range
Cases Versus Extreme Cases
Race does not affect
all cases equally. Notorious serial killers like Ted Bundy or John Wayne
Gacy, both white, are nearly certain to receive the death penalty regardless
of their race. In the most highly aggravated cases, the fact that the
defendant is black is less of a factor pushing a case toward a death
sentence. The same can be said for cases of very low severity: race
is less likely to be a factor in cases where there is little inflammatory
evidence.
But in the "mid-range"
of severity (or aggravation), race plays a very significant role. When
cases were ranked from 1 to 8 in increasing severity, cases in categories
1 (least severe) and 8 (most severe) showed little or no discrimination
against black defendants. But in the middle categories 3 through 7,
the disproportionate treatment of black defendants, as compared to all
other defendants, was quite pronounced. For example, in cases of level
5 severity, 25% of the black defendants received the death penalty,
but only 5% of the other defendants received death, and the difference
between these sentencing rates is 20 percentage points. At level 6 severity,
the difference was 15 percentage points, and at level 4 severity, the
difference in death sentencing rates was 11 percentage points higher
for black defendants. These results are summarized in the graph below.
In other areas of
society, such as employment or housing, racial disparities similar to
those shown in this death penalty study have raised deep concerns and
have prompted civil rights legislation to protect the rights of minorities.
But with the death penalty, this clear evidence of racial bias has gone
uncorrected.
Black
Defendants and the Race of the Victims
Another measure
of race's impact on the death penalty is the combined effect of the
race of the defendant and the race of the victim. In the Philadelphia
study, the racial combination which was most likely to result in a death
sentence was a black defendant with a nonblack victim, regardless of
how severe the murder committed. Black-on-black crimes were less likely
to receive a death sentence, followed by crimes by other defendants,
regardless of the race of their victims.
As noted above,
in cases deemed to be least severe and those found to be most severe,
the connection between race and the likelihood of a death sentence tends
to lessen. But for the bulk of crimes which are in the mid-level of
severity, blacks who kill nonblacks are more likely to receive the death
penalty than blacks who kill blacks, and they have a death sentencing
rate much larger than the rate for defendants of other races who commit
similarly severe murders of black victims.
It is important
to note that these mid-range cases are precisely the ones in which prosecutors
and jurors have the most discretion on seeking and imposing the death
penalty. And when discretion is more prevalent, race may more easily
become the deciding factor in who lives and who dies.
Black defendants,
regardless of their victims' race, are consistently more likely
to receive a death sentence than other defendants, and this holds true
to varying degrees throughout the increasing levels of crime severity.
Similarly, black victim cases are less likely to receive
the death penalty, regardless of the race of the defendant.
Philadelphia
Study: Conclusions
After controlling
for levels of crime severity and the defendant's criminal background,
the average death sentencing rates in Philadelphia were .18 for black
defendants and .13 for other defendants, which amounts to a 38% higher
rate for blacks.
Whichever measures
the researchers employed, the statistics pointed to the same conclusion:
black defendants on average face a distinctly higher risk of receiving
a death sentence than all other similarly situated defendants. The various
independent tests were so thoroughly consistent that they pointed to
race discrimination as the underlying cause. The researchers stated:
"In the face of these results, we consider it implausible that
the estimated disparities are a product of chance or reflect a failure
to control for important omitted case characteristics. . . . In short,
we believe it would be extremely unlikely to observe disparities of
this magnitude and consistency if there were substantial equality in
the treatment of defendants in this system."
