Deterrence and the Death Penalty (2012) / Chapter Skim
Currently Skimming:
4 Panel Studies
4 Panel Studies
Pages 47-74
The Chapter Skim interface presents what we've algorithmically identified as the most significant single chunk of text within every page in the chapter.
Select key terms on the right to highlight them within pages of the chapter.
From page 47... ...
Over this time period, there have been variations in the frequency of death penalty sentences, executions, and the legal availability of the death penalty. With these types of data, the strategy for identifying an effect of the death penalty on homicides has been, roughly speaking, to compare the variation over time in the average homicide rates among states that changed their death penalty sanctions versus those that did not.
|
From page 48... ...
We then discuss the primary challenges to researchers using panel data and methods to inform the question of whether the death penalty affects the homicide rate: the difficulty in measuring changes over time in the relevant sanction policies for homicide and the difficulties in establishing that any changes in homicides that are concurrent with changes in the death penalty are caused by those changes in the death penalty and not vice versa or by other factors that affect both -- such as other sanctions for murder. We conclude with our assessment of the informativeness of the panel research.
|
From page 49... ...
In addition to these fixed effects, some of the researchers also include statespecific linear time trends that allow each state's homicide rate trend to vary (linearly) from the year-to-year national fluctuations.
|
From page 50... ...
estimate 55 different panel data regression models. In 49 of the models, the estimated effect of capital sanctions on homicide is negative and statistically significant; in 4, the estimates are negative and insignificant; and in 2, the estimates are positive and insignificant.
|
From page 51... ...
argue that death penalty sanctions are likely to be correlated with unobserved determinants of homicide, and instead propose using instrumental variables to provide variation in the risk perceptions of potential murderers that is separable from the effects of all of the unobserved factors. The results of and Frakes and Harding reported substantial deterrent effects.
|
From page 52... ...
Figure 4-1 illustrates a partial regression plot with a death penalty sanction measure on the horizontal axis and the homicide rate on the vertical axis (adjusted for state and year fixed effects and typical covariates)
|
From page 53... ...
The regression was run on data for 1984-1998, weighted by state population share, and standard errors were clustered by state. The coefficient of the ordinary least squares line between these two sets of adjusted variables -- and hence the coefficient on the execution measure in the multiple linear regression of homicide rates on the execution measure and all covariates -- is –0.183 (p = 0.173)
|
From page 54... ...
Two mechanisms that could plausibly create associations between changes in death penalty and prison sentence sanctions for homicide are the plea bargaining process, through which the threat of the death penalty may change the likelihood of sentences of different lengths, including life without parole, and the punitiveness of a state's culture, which influences the severity of the capital and noncapital aspects of the sanction regime. None of the studies we reviewed made any use of information on other sanction risks for murder or the ways in which they may be changing over time.
|
From page 55... ...
Consider, for example, the specifications used for variables described as the risk of execution given a death sentence: • the number of executions in the prior year (prior to the current year's homicide rate) ; • the number of executions in the prior year divided by the number of death sentences in the same prior year (or a variant, using a 12-month moving average of these counts for both the numerator and denominator)
|
From page 56... ...
As potential murderers may be attempting to predict the effective sanction regime several or many years into the future, when they might be sentenced or executed, it is particularly unclear what the relevant geographic or time horizon is for obtaining a salient measure. Suppose that when deciding whether to commit a crime, potential murderers weigh the benefits and risks that committing murder may bring them along with the likelihood of those benefits or risks occurring.
|
From page 57... ...
Either implicitly or explicitly, researchers in this field typically make an additional assumption that the risk perceptions of potential murderers are accurate and thus the perceived risks of receiving a death sentence, being executed, or being executed within a particular time period, are equivalent to the objective measures of these risks. The accuracy of this assertion that the risk perceptions of potential murderers are correct is questionable.
|
From page 58... ...
As a consequence, even for well-informed potential murderers living in states with similar sanction regimes, one would expect sanction risk perceptions to evolve along different paths that would depend, among other things, on the size of the state. Perhaps in an environment in which sanction regimes were plausibly stable, the objective risk of execution could be precisely estimated even in small states with low murder rates.
|
From page 59... ...
Thus, using a fixed number of years of lag between those sentenced and those executed means that for many states and years this lag will have an uncertain relationship to the objective risk of execution given a death sentence. The fact that there is a mismatch between the numerator and denominator in the models used is perhaps best illustrated by the many state-year cases in which there are one or more executions the prior year but there were no death sentences imposed 7 years earlier.
|
From page 60... ...
TABLE 4-2 Death Sentences and Removals, by Jurisdiction and Reason for Removal, 1973-2009 60 Total Sentenced Sentence or Under Sentence of Removals to Death, Conviction Sentence Other Death, December 31, Jurisdiction 1973-2009 Executed Died Overturned Commuted Removals 2009 U.S. Total 8,115 1,188 416 2,939 365 34 3,173 Federal 65 3 0 6 1 0 55 Alabama 412 44 31 135 2 0 200 Arizona 286 23 14 110 7 1 131 Arkansas 110 27 3 38 2 0 40 California 927 13 73 142 15 0 684 Colorado 21 1 2 15 1 0 2 Connecticut 13 1 0 2 0 0 10 Delaware 56 14 0 25 0 0 17 Florida 977 68 53 447 18 2 389 Georgia 320 46 16 147 9 1 101 Idaho 42 1 3 21 3 0 14 Illinois 307 12 15 96 156 12 16 Indiana 100 20 4 54 6 2 14 Kansas 12 0 0 3 0 0 9 Kentucky 81 3 6 35 2 0 35 Louisiana 238 27 6 114 7 1 83 Maryland 53 5 3 36 4 0 5 Massachusetts 4 0 0 2 2 0 0
|
From page 61... ...
Mississippi 190 10 5 112 0 3 60 Missouri 182 67 10 52 2 0 51 Montana 15 3 2 6 2 0 2 Nebraska 32 3 4 12 2 0 11 Nevada 147 12 15 36 4 0 80 New Hampshire 1 0 0 0 0 0 1 New Jersey 52 0 3 33 8 8 0 New Mexico 28 1 1 19 5 0 2 New York 10 0 0 10 0 0 0 North Carolina 528 43 21 297 8 0 159 Ohio 401 33 20 168 15 0 165 Oklahoma 350 91 12 165 3 0 79 Oregon 58 2 2 23 0 0 31 Pennsylvania 399 3 24 148 6 0 218 Rhode Island 2 0 0 2 0 0 0 South Carolina 203 42 5 98 3 0 55 South Dakota 5 1 1 1 0 0 2 Tennessee 221 6 15 105 4 2 89 Texas 1,040 447 38 167 56 1 331 continued 61
|
From page 62... ...
TABLE 4-2 Continued 62 Total Sentenced Sentence or Under Sentence of Removals to Death, Conviction Sentence Other Death, December 31, Jurisdiction 1973-2009 Executed Died Overturned Commuted Removals 2009 Utah 27 6 1 9 1 0 10 Virginia 150 105 6 14 11 1 13 Washington 38 4 1 25 0 0 8 Wyoming 12 1 1 9 0 0 1 Percentage 100 14.6 5.1 36.2 4.5 0.4 39.1 NOTE: Some inmates executed since 1977 or currently under sentences of death were sentenced prior to 1977. For those inmates sentenced to death more than once, the numbers are based on the most recent death sentence.
|
From page 63... ...
(Of course, one can only speculate about which, if any, of these variables is salient for potential murderers.) These many complications make clear that even with a concerted effort by dedicated researchers to assemble and analyze relevant data on death sentences and executions, assessment of the actual and changing objective risk of execution that faces a potential murderer is a daunting challenge.
|
From page 64... ...
that the sanction regimes of adjacent states do not have any bearing on the effect of the death penalty in a particular state. We begin with a brief discussion of the benefits of random assignment.
|
From page 65... ...
This approach identifies a causal effect only if there are no other factors besides the death penalty causing homicide rates to change differently in states that do and do not experience changes in death penalty sanctions. Many such factors may well exist -- such as changes in economic conditions, crime rates, public perceptions or political regimes -- and there is no reason to believe that these variables are fixed over time or across states.
|
From page 66... ...
in other aspects of the sanction regime for murder, then the required assumption is violated, and those states cannot provide the missing counterfactual information for states that do experience changes in the death penalty. A related concern is that while death penalty sanctions may be affecting the homicide rate, the homicide rate may also be affecting death penalty sanctions and statutes.
|
From page 67... ...
Compounding the problem, even less is known about factors that are associated with death-penalty-relatedchanges in the sanction regime for murder, or more relevantly, changes in perceptions of sanction risks. As noted above, factors contributing to changes in the legal status of the death penalty or the intensity with which the death penalty is applied could include economic, crime, or political changes that may also have direct consequences for the homicide rate.
|
From page 68... ...
The problem, however, is finding variables that are related to the sanction regime but not directly related to homicide rates. In general, the committee finds that the instruments proposed in the research are not credible and, as a result, this identification strategy has thus far failed to overcome the challenges to identifying a causal effect of the death penalty on homicide rates.5 Homogeneity Still another assumption of the panel regression model in Equation (4-1)
|
From page 69... ...
These models assume that every possible death-penalty-intensity level would have the same effect on homicide rates in every state and year if it was present in that state and year, regardless of the prior sanction regime, a state's history with the death penalty, or any other factor. Although this homogeneity assumption is commonly invoked in regression models, no support is offered for it in studies of the death penalty, and on its face it appears unlikely to hold.
|
From page 70... ...
The sanction regime is insufficiently specified and the measures of the intensity with which the death penalty is applied are flawed. No connection has been established between these measures and the perceived sanction risks of potential murderers.
|
From page 71... ...
In addition, the key assumption that the death penalty sanction is independent of other unobserved factors that might influence homicide rates seems untenable. For these reasons, the fixed effects models are no more informative about the effect of the death penalty on homicide rates than other types of model.
|
From page 72... ...
. Getting off death row: Commuted sentences and the deterrent effect of capital punishment.
|
From page 73... ...
. Deterrence versus brutalization: Capital punishment's differing impacts among states.
|
Key Terms
This material may be derived from roughly machine-read images, and so is provided only to facilitate research.
More
information on Chapter Skim is available.