Since the death penalty was brought back into the American legal system in 1976, it has triggered heated debates. No longer unconstitutional, states were given the power to choose if they wanted to implement the death penalty as a form of punishment. The debate seems to hinge on the idea of whether the death penalty actually deters crime or not. The idea behind this research is to understand capital punishment’s effect on crime by comparing a specific unit of analysis. In this case, the unit of analysis will be the states, ideally it would be best suited to compare all of the states but realistically the key number of comparison is 28 states.
This research is significant because it will attempt to show if the death penalty does deter crime. The United States is the only country in the Western world that still uses capital punishment as a sentence. Research proving either side of the debate to be true will have a considerable impact on not only those affected first hand by the punishment but also the nation’s view as well as how the rest of the world looks at us. One of the main purposes of the research is to show a strong significant correlation one way or the other between the death penalty and homicide rates.
The scope of the research is narrowed down to compensate crime relative to individual states, which will give a unique perspective of how the specific state’s homicide rates compare to each other based off whether or not they impose the death penalty as a form of punishment. By using specific states as the unit of analysis it is easier to control the perspective of the task at hand while keeping the information processing at a simplified level. It is the hope that this research accomplishes the main goal of finding a correlation between the death penalty and various crime rates in different states in order to show if the punishment does or does not have a deterrent effect.
The death penalty is only instituted in America because the public feels that it is a necessary punishment to capital murder. In the late 90’s public opinion polls showed an overwhelming majority of Americans favored the death penalty with some polls showing almost 80% (Radelet, 1). One would most likely be safe to assume that this is an accurate number, because if the polls were lower or even closer to being 50-50, then perhaps the death penalty wouldn’t be implemented anymore. Well, give the public the option of life in prison without the possibility of parole, and now all of a sudden those high numbers significantly drop to almost 50% (Radelet, 3). Interestingly enough, life in prison without the possibility of parole is a very real choice that can be made.
It is worth noting that these people, who are advocates of the death penalty due to its deterrent effect, seem to rely on a study widely criticized for its methodological procedures and faulty analysis. Ehrlich’s research done from the early 30’s into the late 60’s found that there was not only a significant deterrent effect but that for “an additional execution per year over the period in question may have resulted, on average, in seven to eight fewer murders” ( Bailey, 1311). This data has been widely criticized by numerous researchers all of whom could not replicate his findings. Bailey cites such issues as his misleading probabilities and the problem of generalizing his findings to a whole nation when clearly states differ quite differently on who gets executed and who doesn’t (Bailey, 1311).
Going off the notion that states differ so vastly when it comes capital punishment and number of executions, it would be obvious that to test the homicide rates to those states with and without the death penalty, one could calculate whether it serves as deterrence in those states. Obviously this is a lot simpler said than done, but numerous studies to that general concept have been conducted and all in all show that there is no deterrent effect when it comes to homicide rates in comparison to whether or not the death penalty is imposed ( Fisher, 53). Not be completely one sided against the work done by Ehrlich, these studies also have their problems when it comes to the methodological techniques used. There are many variables that cannot be accounted for and it is very hard to account and have controls for all these problems that may be encountered. So, as will be discussed later, by having more variables there are less extraneous variables left to account for.
Although the study we will be conducting deals with state to state unit of analysis, one of the studies down that should be noted due to the impeccable methodology used is the study done by Cochran Chamlin, and Seth in the late 90’s. Instead of comparing different states which is hard because of the problem of outside variables and differences in each state, Cochran compared just Oklahoma before and after the death penalty was re-implemented (Fisher, 53). Here, once again the findings show that there is no significant deterrent effect.
In a 2001 study conducted by Rankin, a more unique approach to the relationship between violent crime and homicide rates are examined. In this study, contrary to other research done in the field at the time, Rankin found that there is a nonlinear relationship between support for capital punishment and the rate of violent crime. This aspect is less studied in the field because collecting data on attitudes at particular points in time can be difficult to do in a reliable but quick manner. But, from this study one would be able to conclude that as crime rates go up, so do people’s need for implementation of the death penalty and so that brings into play a whole new variable of attitude that isn’t mentioned.
It is evident that whether or not the death penalty serves as a deterrent to homicide rates is blurry at best due to the lack of being able to control for variables and to comparably match states based on the same criteria before analyzing them. While some states have the death penalty still on the books, their execution rates may be much lower or the population may be very different and so individually it is hard to find a general and clear statement that the death penalty does anything for homicide. So, as some research in the past has done it is necessary to look at general patterns along with homicide rates when doing a broader study that include numerous states.
One such area studied has to do with religion and state implementation of capital punishment. In Bjarnason’s 2004 research he conducted a study to examine the effects of numerous attitudes in relationship to state death penalty implementation including Catholic Parishioners and their relationship with state death penalty laws. Surprisingly, it was found that although the Catholic Church may be a strong advocate against the death penalty, there is really no significant difference between Catholic support and non-catholic support on implementation of the death penalty.
In a study done in 2004, Jacobs and Carmichael found that there was a significant difference in whether or not the state with the death penalty has a higher or lower rate of conservative churches. They found using a two-equation count models that along with violent crime rate and minority rate, the higher the conservative church population is, the more likely the state is to employ the death penalty. Conservatives is a word possibility used to broadly in this study, but their hypothesis was that where “conservative” values were the strongest there would be more demand for death sentences goes completely against Bjarnason’s hypothesis and conclusion that there was no difference.
Little attention has been paid to areas outside of religion, violent crime and homicide rates, but we think that it is important to explore aspects that might seem more trivial but could bring about new discussion and answers. There is no research surrounding the topic of how property crime is related to the death penalty and although that may seem like two different topics completely, when arguing the broken window theory it all comes back to that first broken window per se. When talking about states use of death penalty implementation, how far back the problems start could answer a lot of questions early on and it would be not only beneficial but perhaps a start to fixing something before it begins.
The following statements are the hypotheses for this research project:
- States that impose the death penalty have different homicide rates than states that do not.
- States with a higher percentage of Catholic’s in the population have different homicide rates than states with a smaller percentage.
- States that impose the death penalty have different violent crime rates than states that do not impose the death penalty.
- States that impose the death penalty have different property crime rates than states that do not impose the death penalty.
- States with higher execution rates have different homicide rates than states with lower execution rates.
In order to collect all the necessary data so that our analysis was most accurate, a highly quantitative method was used. Due to having more than one dependent variable, we had to gain our statistical data from multiple sources before putting it all together into one. Our first concern when collecting data was finding data specific to each state. So, the main goal first through dependable on line sources was getting the current population of all 50 states.
This was necessary to then chose 14 states that matched population wise with the 14 states that didn’t impose the death penalty legally. After we had gained 28 states that matched population wise, we were able to do a highly thorough literature review on the topic of capital punishment and deterrent effect to see what the most current studies had been doing and what results they were finding. After finding that there was most definitely a current trend on if a deterrent effect was really in play, we were able to narrow down our own variables that we thought could further the current research along with some new variables that we thought could open new doors.
After the literature review was completed it was decided that the most productive and least time consuming study for our purposes would be by using states for our unit of analysis. With this in mind, it was than necessary to come up with our variables that would be measured and correlated against each other on a state level with the two different main independent variables being whether or not the state legally implemented the death penalty. So, using our two independent variables and five dependent variables seen in our hypothesis mentioned above, we were able to analyze the data to come up with our results.
Through this process there were a few problems that did arise when collecting our data. Finding the most current state data for every variable was difficult at times and we had to make sure the years were the same to exclude other extraneous variables. Also, once finding the most recent data other resources had to be used to make sure that the data was accurate and reliable.
Another weakness related to the two just mentioned was that when getting our matching states that did impose the death penalty the only thing we used to match up the states was population. We didn’t have the resources to really take into account other variables that would more evenly match states such as geographic location or the types of people residing in each state.
On the other hand, there was definitely some strength in our decision to make our unit of analysis states. This left little variability for other extraneous factors because the 28 states were spread all across the country. So, the findings were able to be generalized to the whole country, not a specific region which was a problem we were able to avoid after understanding the literature and studies that had already been done on the topic. Our methods used to analyze the date were only as complex as what we had on hand and so our findings can only go so far on a university level with limited resources and that is important when discussing our findings.
Results and Findings
The purpose of this research was to find a strong significance between the death penalty and crime. The database of 28 states (the unit of analysis) was broken down into a frequency distribution of 14 states that implement the death penalty and 14 states that do not (the 14 states that impose the death penalty were chosen from the field of 36 based on comparable populations to the 14 states that do not implement the death penalty). The six variables that were used in correlations were population, homicide rates, violent crimes rates, execution rates, property crime rates, and percentage of Catholic population.
Research projects need to ultimately answer questions and this project is no different. We asked a set of very specific questions before doing an extensive literature review and came up with five hypotheses. The questions, first presented in the introduction, are listed for easier comprehension here:
- Does the death penalty deter crime?
- Do states that implement the death penalty have less crime than states that do not use the death penalty?
- Is there a correlation between the death penalty and homicide rates?
Once we determined what this research was intending to find, we set out and found a plethora of peer reviewed journal articles that narrowed down the questions and off of these we created our five hypotheses (seen in Literature Review). After the hypotheses were established, it was imperative to collect all of the updated data (as explained in Research Methods) necessary to complete the research and answer or research questions. For each correlation based on the specific hypothesis, the independent and dependent variables changed. The following list shows the five hypotheses correlations, in order of the previous stated hypotheses, and labeled variables:
1. Independent: Population Dependent: Homicide Rates
2. Independent: Percentage of Catholics Dependent: Homicide Rates
3. Independent: Population Dependent: Violent Crime Rates
4. Independent: Population Dependent: Property Crime Rates
5. Independent: Execution Rates Dependent: Homicide Rates
The next step was to create a codebook and enter our collected information into SPSS. Attached are the results of the bivariate correlation analysis done to determine the hypotheses. Bivariate correlation analysis was used instead of the Chi-square test because our statistics were based on a scale measurement rather than a nominal or ordinal measurement. The findings at first glance are not clear-cut because it shows that for states that implement the death penalty there were no strong significant correlations, or the correlation pearson score, P score, was .05 and was adjusted accordingly. For states that do not have the death penalty as a form of capital punishment, there was only a strong correlation between violent crime rates and the population.
However, we set out to find if states that do not implement the death penalty had different crime rates, specifically homicide rates, than states that used the death penalty. The results show and prove three out of the five hypotheses. There is an established significant difference between states that impose the death penalty and those that not in the following categories: states that impose the death penalty have higher homicide rates than states that do not impose it, states with a higher Catholic population percentage have less homicide than states with a small Catholic population percentage, and violent crime rates are higher in states that impose the death penalty. Since it is impossible to correlate execution rates in states that do not implement the death penalty, we can neither accept nor reject the hypotheses that states with higher execution rates have different homicide rates than states with lower execution rates.
However, we can conclude that the hypothesis that states that impose the death penalty have different property crime rates than states that do not impose the death penalty is not true. There is too small a difference between the two correlations of the groups of specific states. We determined that if the difference between the two correlations was less than .05 that there was not a strong difference between the two and this statistical significance disproves the hypothesis.
Back to the main question of does the death penalty deter crime? We believe that we amassed enough literature, statistics, information and research, to determine an answer for not only this question but the other research questions we established as well. Based off our data, we can say it is safe to say that the death penalty does not deter crime, specifically violent crime and especially homicide. A pleasant surprise returned was the conclusion that if states have a higher percentage of Catholics in their populations they have lower homicide rates. However, since we rejected our property crime hypothesis we believe that states that implement the death penalty and states that do not are both as susceptible to property crime. Our findings are solid statistical relationships based on facts and reliable SPSS software.
Analysis and Discussion
The main point of this research project was to find out if there was a significant difference between crime in states that impose the death penalty and states that do not. The simple way to do that would be to look at states that do not impose the death penalty and their all of their crime rates and states that do impose the death penalty and all of their crime rates. To take the research further, the researchers wanted to figure our more about the death penalty and decided to use other variables that the use of the death penalty might affect. It is important to see all sides to the argument and to be able to identify if the death penalty deters all crime, deters murder, or does not work at all. Understanding this concept was a difficult task and the use of the SPSS program benefited the research.
To use SPSS correctly, a codebook had to be created in order to ensure academic integrity and accuracy. The first variable used was population. We created a database of states, precisely using the 14 states that do not implement the death penalty and conversely used 14 states that matched those states the closest in population that did implement the death penalty. Using a scale measurement, we put state populations on a four-point scale between 500,000 and 20,000,000 (for further information see attached Code Book). This process was used five more times using the five variables selected for correlating, homicide rates, percent Catholic, violent crime rates, property crime rates, and execution rates (see attached Frequencies list). All six of the variables frequency charts came out correct.
The research done was a long process that took time, devotion, patience and hard work to complete and do right. In conclusion, we answered all of our main questions we came into this project looking to answer and learned some more along the way. Based on our research and careful statistical analysis we can conclude that there are many differences in crime between states that impose the death penalty and states that do not.
Bailey, William C. (June 1980). Deterrence and the Celerity of the Death Penalty: A
Neglected Question in Deterrence Research. Social Forces 58, (4), 1308-1332.
Bjarnason, Thoroddur and Michael R. Welch. (2004). Father Knows Best: Parishes,
Priests, and American Catholic Parishioners’ Attitudes Toward Capital
Punishment. The Journal for the Scientific Study of Religion 43, (1), 103-118.
Bottum, Joseph. (2005). Christians and the Death Penalty. First Things: A Monthly
Journal of Religion & Public Life. 17-21.
Cholbi, M. (2006). Race, Capital Punishment, and the Cost of Murder. Philosophical
Studies 127, 255-282.
Clement, Douglas. (June 2002). A Punishing Debate. The Region. 12-15, 40-42.
Decker, Scott H. and Carol W. Kohfield. (1990). The Deterrent Effect of Capital
Punishment in the Five Most Active Execution States: A Time Series Analysis. Criminal Justice Review 15, (2), 173-191.
Hagan, Frank E. (2006). Research Methods in Criminal Justice and Criminology.
Boston, Allyn and Bacon.
Fisher, Patrick and Travis Pratt. (Mar. 2006). Political Culture and the Death Penalty.
Criminal Justice Policy Review 17, (1), 48-60.
Jacobs, David and Jason T. Carmichael. (Sep. 2004). Ideology, Social Threat, and the
Death Sentence: Capital Sentences across Time and Space. Social Forces 83,(1), 249-278.
Radelet, Michael L. and Ronald L. Akers. (1996). Policy and Perspectives Deterrence
and the Death Penalty: The Views of the Experts. The Journal of Criminal Law & Criminology 87, (1), 1-14.
Rankin, Joseph H. (Sep.1979). Changing Attitudes Toward Capital Punishment. Social
Forces 58, (1), 194-208.
Whitman, Claudia. (2001). The Death Penalty as the Will of the People. Peace Review
13, (4), 519-523.