Does Death Penalty Deter Killers? No Clear Answer

When the current era of American executions opened in 1977, the firing squad bullets that tore into Gary Gilmore’s chest seemed to stop more than just him.

Gilmore’s execution in Utah, the first in the country after a 10-year moratorium, was followed by a brief but sharp drop in the national homicide rate.

Death penalty advocates claimed that heavy publicity about Gilmore’s execution had deterred would-be killers by making them think that what happened to Gilmore could happen to them.

But death penalty opponents had another explanation. They noted that Gilmore’s execution coincided with one of the worst blizzards ever to hit the eastern United States. It was the harsh weather, they said, and not the deterrent effect of the death penalty, that had kept killers at bay.

The existence of competing, plausible theories such as these for periodic dips and rises in the homicide rate helps explain why the debate over the deterrent effect of capital punishment remains unresolved. And that debate has been revived as California prepares to impose the death penalty April 3 on Robert Alton Harris. His execution would be California’s first in 23 years.

Common sense may suggest that executing people convicted of premeditated murder would deter other would-be murderers. But social scientists, despite years of trying, haven’t been able to prove it. In fact, they have failed to turn up any reliable statistical evidence that it does.

As philosopher and death penalty expert Hugo Adam Bedau of Tufts University said, in summarizing the research: “No reliable scientific investigations support the common sense inference that since the death penalty is more severe than long-term imprisonment, the death penalty must be a better deterrent.”

Criminologist James Q. Wilson agreed, but cautioned: “One could easily imagine a society in which the death penalty would have a deterrent effect, that is, one in which people were executed immediately . . . and invariably.”

Even at the peak of executions in this country, in the 1930s, swiftness and certainty were not the rule: There was only one execution for every 100 criminal homicides, experts said.

Now, nationwide, the ratio is more like one in a thousand. Nationally, there are about 20,000 homicides a year.

“Maybe if we had 10,000 executions” there would be an observable deterrent effect, said Franklin R. Zimring, a UC Berkeley law professor and capital punishment expert. “I don’t know. But we don’t live in that kind of a world. We never have.”

With the odds against a murderer being executed so strong, it is difficult to show that the death penalty has a greater deterrent effect than the threat of long imprisonment, the social scientists say.

“I think it is a subject which social scientists simply cannot answer,” Wilson said, “because the statistical tools we have at our disposal are imperfect.”

The tools are not sensitive enough, he said, to factor out all the other possible causes for ups and downs in the first-degree murder rate.

Politicians like Gov. George Deukmejian have been arguing for years that executing murderers will save innocent lives by discouraging other would-be killers. A Gallup Poll in 1986 found that 61% of Americans agree.

But social scientists say this is largely a matter of conjecture and, as such, is subject to unending debate.

Obviously, the threat of being put to death did not deter the more than 2,200 murderers now on Death Row across the nation. But why not? Didn’t they have the common sense to avoid the threat of the harshest of all punishments?

New York University law professor Anthony G. Amsterdam, a prominent and passionate death penalty foe, has said he thinks not. “Those who are sufficiently dissocialized to murder are not responding to the world in the way that we are, and we simply cannot ‘intuit’ their thinking processes from ours,” he has said.

Therefore, he has argued, it is debatable whether capital punishment will deter them.

The typical murderer sentenced to death in this country has robbed a convenience store, tied up one or more employees, then shot them to death, experts said.

Believers in the deterrence value of capital punishment argue, in part, that such murderers were not dissuaded because they considered it unlikely the death penalty would be imposed even if they were caught.

They also argue that there is evidence that at least some would-be killers have been deterred. Los Angeles County Deputy Dist. Atty. Sterling Norris, who has prosecuted 10 death penalty cases and has put four men on the current Death Row, said his office once documented scores of incidents in which victims said they overheard suspects decide not to kill them for fear of the death penalty.

“It doesn’t deter everybody. It never will,” Norris said. “But it deters some people.”

Texas Atty. Gen. Jim Mattox said in 1988 that he had interviewed nearly all of the people executed in that state since the U.S. Supreme Court-imposed moratorium on the death penalty ended in 1976, and that the possibility of being sentenced to death never crossed their minds. Mattox, however, remains a strong supporter of the death penalty.

Texas has executed 33 34 murderers--far more than any other state. It is followed by Florida with 20 executions and Louisiana with 18.

A look at homicide rates in those states shows the uncertain effect of the death penalty. In Florida, the homicide rate soared in the years immediately after the state resumed in 1979, then fell and has held fairly constant. In Texas, the rate soared before executions began in 1982, then generally fell. In Louisiana, the rates have generally fallen--with wide fluctuations from year to year--since executions began in 1983.

But the rates in all of the those states have generally followed national patterns, which saw homicide rates surge in the late 1970s and decline in fits and starts for much of the last decade before rising again, experts said.

Homicide rates in Florida, Louisiana and Texas have been consistently higher than in California, which has had an unused death penalty on its books since 1977, and New York, which does not have the death penalty.

For most social scientists, these statistics simply underscore the possibility that factors in addition to the death penalty influence fluctuations in homicide rates.

Despite widespread popular support for capital punishment, the United States, as a society, has been reluctant to impose it. Even before the Supreme Court stepped in with its temporary ban on capital punishment in 1972, the number of executions had dwindled for decades.

Although 37 states now have death penalty statutes, the chances of a killer being executed remain slim for a variety of reasons. For one, some people who kill are not caught. Others who are caught are charged with unpremeditated or provoked killings that are generally not punishable by death. Some who are charged with premeditated murders are not convicted, or are convicted but not sentenced to death. And for those who are sentenced to die, appeals usually drag on for years.

Because validating the theory of deterrence necessarily involves exploring killers’ states of mind, social scientists attempt to measure the deterrent factor indirectly--inferring it from other evidence to try to understand the behavior of killers as a group.

One of the pioneers of deterrence research was Thorsten Sellin, a University of Pennsylvania sociologist who was hired in the 1950s by a lawyer’s group trying to decide whether to include the death penalty in a proposed model penal code for the states.

In an effort to determine whether the death penalty deterred killings, Sellin compared homicide rates in states with the death penalty and states without it. He tried to control for social and economic factors that might have influenced homicide rates by comparing states that were next to each other and had similar urban populations, per capita incomes and degrees of unemployment.

He found that homicide rates were about the same in the states that had the death penalty and in the states that didn’t.

He also looked at states that abolished the death penalty or adopted it, and found that adoption or abolition “exercises no influence on the extent or fluctuating rates of capital crimes.”

“The inevitable conclusion,” he wrote, “is that executions have no discernible effect on homicide death rates.”

Critics of his methods said the states he compared were not similar enough, and that many other factors could have obscured the impact of the death penalty.

But his conclusions were not seriously challenged until 1975, when Isaac Ehrlich, then of the University of Chicago, applied a more mathematically sophisticated statistical model--of the kind economists use.

Ehrlich looked for evidence that the “supply” of murders could be reduced by the negative social “demand” for them, as manifested by the number of executions.

Assuming that killers behaved rationally, Ehrlich figured that the death penalty might have added an incentive for some of them to murder more witnesses in an effort to escape detection. But, overall, his analysis suggested that each execution might have saved as many as eight innocent victims from being killed.

His findings caused a stir, and a panel of the National Academy of Sciences was named to study his methods. The panel found them so wanting that it declared that his analysis “provides no useful evidence on the deterrent effect of capital punishment” which is “definitely not a settled matter.”

One of the major criticisms of Ehrlich’s work--and that of many others in the field--is that it relies on unreliable data for the most basic of its concerns: the number of murders that conceivably could be deterred by the death penalty.

Only certain first-degree murders are punishable by death. But statistics on first-degree murders are simply not available from law enforcement or public health sources. Instead, researchers have had to make do with statistics on homicides in general in attempting to analyze the effect of the death penalty. These general statistics are tainted, experts say, because many homicides are manslaughters that are not punishable by death because they are either unpremeditated or provoked.

The most compelling criticism of this kind of research, however, is that, as a practical matter, it is impossible to isolate the effect of the death penalty as a deterrent from that of many other social and personal factors that may influence decisions to kill.

Ehrlich and others have tried by mathematically accounting for the effects of some of the other factors. They have then attributed what is left over to the death penalty’s deterrent effect.

But other researchers have argued that this method has a built-in bias in favor of the deterrent effect.

As Zimring, of UC Berkeley, put it, when you structure a study that way, “You’re going to come out with any answer you want to. . . . Any statistical residual that’s left you attribute to whatever you’re investigating.”

MURDERS AND NON-NEGLIGENT MANSLAUGHTERS

Here are figures on homicide rates per 100,000 population for California, which has had the death penalty since 1977 but has executed no one ; for New York, which does not have the death penalty ; and for Texas, Florida and Louisiana, which have executed more inmates than any other states since a moratorium imposed on capital punishment by the U.S. Supreme Court was lifted in 1976. Texas has executed 34; Florida, 20; Louisiana, 18.

YEAR CALIFORNIA NEW YORK TEXAS FLORIDA LOUISIANA 1975 10.4 11.0 13.4 13.5 12.6 1976 10.3 10.9 12.2 10.7 13.2 1977 11.5 10.7 13.3 10.2 15.5 1978 11.7 10.3 14.2 11.0 15.8 1979 13.0 11.9 16.7 12.2* 16.9 1980 14.5 12.7 16.9 14.5 15.7 1981 13.0 12.3 16.6 15.0 15.6 1982 11.2 11.4 16.1* 13.5 16.0 1983 10.5 11.1 14.2 11.2 14.2* 1984 10.6 10.1 13.1 11.5 12.9 1985 10.5 9.5 13.0 11.4 10.9 1986 11.3 10.7 13.5 11.7 12.8 1987 10.6 11.3 11.7 11.4 11.1 1988 10.4 12.5 12.1 11.4 12.1

*Denotes the year that executions were resumed in the state.

Source: FBI Uniform Crime Reports, which do not differentiate between killings that are punishable by death and those that are not.

Compiled by Times editorial researcher Michael Meyers.