What Determines a Person Support or Opposition to Death Penalty - Report Example

What Determines a Person Support or Opposition to Death Penalty? The Case For and Against Death Penalty Capital punishment has been a common practicein Western civilization for over 2,000 years. It has been used to punish individuals whose activities compromise the peace and security of the society. Over the years, however, its use and abuse has been a subject of numerous debates and a topic of controversy in the United States as most states are now passing laws to eradicate it. Those who oppose it claim that the death penalty system is seriously flawed (Randall 2000). To prove this claim, a study entitled The Justice Project reviewed 4,578 capital cases that were appealed between 1973 to 1995. 313 of these cases resulted in execution but the study showed that as much as 68%, or 7 out of 10 cases, could have been reversed. According to the researchers, “capital trials produce so many mistakes1 that it takes three judicial inspections to catch them—leaving grave doubt whether we do catch them all. After state courts threw out 47% of death sentences due to serious flaws, a later federal review found ‘serious error—error undermining the reliability of the outcome—in 40% of the remaining sentences” (as quoted in Randall 2000). In response to this study, Illinois governor George Ryan pardoned four death row prisoners and commuted the other 163 cases into life imprisonment. This move has garnered widespread criticism from death penalty supporters who accused Ryan for seeking public favor redemption for his corrupt activities. Meanwhile, pro-life forces have nominated the ex-governor for the Nobel Peace Prize as recognition. In another case, Attorney General John Ashcroft used death penalty in order to secure the communities terrorized by a mysterious sniper. When suspects were arrested, Ashcroft steered prosecution away from Maryland where the death penalty was under attack to Virginia, where there is a strong support base for capital punishment. These two cases are just two of the major events that demonstrate the debates for and against capital punishment. There are many others that support the two opposing positions. There are also a number of researches aiming to understand why the death penalty is important, or why it should be abolished instead. This study does not create a position for or against capital punishment. Instead, it seeks to understand how wide a support the practice receives from Americans, as well as to determine the characteristics of individuals who support or object to capital punishment. This is important to know because such information can affect policy concerning death penalty. For example, in the 2009 results of the annual Gallup Crime Survey, they discovered that atleast 65% of Americans continue to support death penalty (Newport 2009). What are the characteristics of these Americans who support death penalty? This is the main question for this research. It is important for policy makers and politicians to know the answer to this question because their own view on capital punishment can affect the support base for future policies and laws they may propose. To expound on the main research question, the following null hypotheses have been formulated: 1. A person’s sex/gender does not affect his or her decision to support or oppose death penalty. 2. Age is not a determinant to the support or opposition to capital punishment. 3. Individuals with higher incomes do not demonstrate greater support to death penalty compared to those with lower income levels. Research Design This study utilizes secondary data from the General Social Survey (GSS) which can be found on this site: http://www3.norc.org/GSS+Website/. On its website, GSS is described as a “major teaching tool…[it] takes the pulse of America, and is a unique and valuable resource” (The National Data Program for the Sciences 2011). Aside from the U.S. Census, the GSS is a frequent source of data for a number of social science courses. The GSS monitors the change in attitudes and opinions of the American society regarding various issues such as capital punishment, reproductive health, military spending, among many others. Access to raw data is free and can be downloaded in SPSS and STATA format, thereby allowing students to test variables and understand how each are related. For this study, SPSS will be used to determine which among the three independent variables - respondent’s sex (coded as sex), age (age), and income (rsincome06) – determines a person’s decision to support or oppose the application of death penalty for persons convicted of murder (cappun). Descriptive statistics such as frequencies, means and crosstabs will be utilized to further make sense of the available data. Levels of Measurement Alan Bryman says that there are three common levels of measurements: nominal, ordinal and interval data. They are referred to as “levels” because they offer various kinds of information and will enable the researcher to use different statistical tests in order to extract connections between variables. The higher level the variable is, the more powerful the statistics (and connection) will be. The lowest level of measurement from the statistical point of view, is the nominal scale. As the name implies, this scale is typically used to “name” or categorize data with no real order, and measurable distance between the categories. In research, items answerable by yes or no are nominal level variables. Moreover, sex is a nominal level variable because there is no implicit order between males and females (though of course, there is a socially constructed difference between these two, in terms of statistical value, one male is equal to one female). Another nominal level variables in this research is opposition or support for capital punishment (CAPPUN). Nominal level data can be analyzed using non-parametric statistics such as chi-square or the mode. Ordinals are regarded as a higher level of measurement compared to nominal variables because they imply order (Bryman 1988). The simplest example of the ordinal level data is ranking. When a researcher asks a respondent to rank services based on how effective they are, one is basically creating an ordinal scale of preference. The first in the list is typically more preferred, but the researcher will not know the exact difference between the first and second or even the fifth in the list. Just because something is ranked 5th, it doesn’t mean that it is five times less preferred than the others, instead, it is the 5th most preferred. In the case stated wherein a respondent is asked to rank services, top 1 and 2 may differ by 2 point while rank 2 and 3 can have a difference of 5 points. In this research, income is considered as an ordinal level data (instead of the usual interval data) because the categories used have relative positional distances. For example, the range $1,000 to $2,999 is lower compared to $3,000 to 3,9999 but the exact differences between these two categories is not clearly defined. Some of the respondents may have answer $3,001, $3,005, $3,100, but these items will be collated in just one category. Ordinal level variables enable researchers to make use of more statistics including non-parametric ANOVA, rank order correlation and median/mode as measures of tendency. Variables using the standard Likert scale are interval-level. Like the ordinal level variable, interval-level variables may be ranked and ordered, but the main difference is that the degree of differences are always the same. Because of its unique features, interval level has greater flexibility and can utilize more powerful statistical procedures including regression, correlation, ANOVA and many others. . In the case of this research, age maybe regarded as interval-level, but because of an additional characteristic, it is more appropriately referred to as ratio-level variable. In this data set, age is regarded as a ratio-level variable because it is arranged in order, and the degree of differences is always the same (Bryman 1988). For example, we know that a person aged 60 years old is older by 10 years to a person aged 50, and that a person aged 40 is younger by 10 years from someone who is aged 50. Aside from the two characteristics mentioned, age is a ratio-level variable because it has an absolute zero. To better understand this, we can compare age (ratio-level) with temperature which is an interval level data. A person aged 0 is someone who is below 1 year old, or is not yet born. Meanwhile, while there is such a thing as 0°C this value does not mean that there is no temperature, instead, there is just the absence of heat. In fact, when 0°C is converted to the Farenheit scale, it would be equivalent to -32° which means that the point of “zero” in interval scale variables are relative whereas in ratio level variables have an absolute zero. Describing the Characteristics of the Sample The available GSS dataset had involved 1,500 respondents in the sample. Figure 1 below shows that majority (53%) of the respondents are females and that there is no real ranking between the two sexes (male and females), and they are given equal weight during data gathering. Sex in this case, is used only to categorize respondents. Other nominal variables for this study includes political affiliation, race and religious preferences. Figure 1: Respondents Sex Income, in the case of this study is regarded as an ordinal data. Using table 2 below, one can gain better understanding why it is categorized as such, instead as an interval level variable. It is clear from the table that more than half (54.6%) of the respondent earned more than $25,000 and above, which means that the respondents for this survey are of the middle to upper class. Because the data for the income is in the ordinal level, it is impossible to say how many are earning $25,000, $30,000 or more. Moreover, when comparing two income ranges such as, we only know taht $1,000 to 2,999 have a lot lesser earnings than those who reported their income at $3,000 to 3,999. Table 1: Frequency distribution of respondents by income Frequency Valid Percent Cumulative Percent Valid LT $1000 18 1.9 1.9 $1000 TO 2999 21 2.2 4.1 $3000 TO 3999 10 1.0 5.1 $4000 TO 4999 19 2.0 7.1 $5000 TO 5999 11 1.1 8.2 $6000 TO 6999 14 1.5 9.7 $7000 TO 7999 7 .7 10.4 $8000 TO 9999 21 2.2 12.6 $10000 - 14999 78 8.1 20.8 $15000 - 19999 74 7.7 28.5 $20000 - 24999 109 11.4 39.8 $25000 OR MORE 524 54.6 94.5 REFUSED 53 5.5 100.0 Total 959 100.0 It could happen, however, that most of those who gave $1,000 to 2,999 are earning $2,999 while those who reported their earnings a $3,000 to 3,999 are only earning $3,000 which means that the mean difference is around $1 only (or it could happen that the mean difference is $100 or more). What we do know from table 2 is that $1,000 to 2,999 is less than $3,000 to 3,999 and that there are more people who are earning the lesser income, according to the GSS. About 5.5% of the respondents refused to give information about their incomes while 541 individuals have missing responses for this item. In terms of age, descriptive statistics show that the youngest respondent for the GSS was aged 18 while the oldest was 89 years old. Most of the respondents (79%) were aged 60 and below, with the mean age at 46.22 years. This means that a majority of the respondents were not yet of old age and are typically in the middle ages (perhaps working, and with their own families). Crosstabulating Data At the beginning of this paper, we have already identified three independent variable and one dependent variable. To address the hypotheses formulated in the earlier part of this paper, crosstabulation and chi-square tests will be performed. Hypothesis 1: A person’s sex/gender does not affect his or her decision to support or oppose death penalty. Table 2 below is a crosstabulation between the respondent’s sex versus their decision to oppose or support capital punishment for individuals convicted of murder. It is clear that in the case of the GSS, more males support applying the death penalty to those convicted of murder while there are more females who oppose this move. How strong is this association? Using Pearson’s Chi-Square test, we find Χ2=12.079 and p=.001. Testing the hypothesis at 95% confidence level, we can say that the correlation between the two variables, sex/gender and the decision to support or oppose death penalty for murder is highly significant. Various literature also confirms the finding that there are more women who oppose death penalty whereas there are more men who support it (Stack 2004; Jones 2003). This means that sex is a strong determinant in the support or opposition of the death penalty. Table 2: Crosstabulation, Respondents sex and decision to support or oppose capital punishment FAVOR OR OPPOSE DEATH PENALTY FOR MURDER Total FAVOR OPPOSE MALE Count 237 77 314 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 51.3% 36.8% 46.8% FEMALE Count 225 132 357 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 48.7% 63.2% 53.2% Total Count 462 209 671 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 100.0% 100.0% 100.0% Hypothesis 2: Age is not a determinant to the support or opposition to capital punishment Testing this hypothesis at α=.05, we have computed Χ2= 71.360 and p=.466. In this case, we can say that the null hypothesis has to be accepted and age cannot be considered as the determinant to the support or opposition to capital punishment. Hypothesis 3: Individuals with higher incomes do not demonstrate greater support to death penalty compared to those with lower income levels. In table 3, we can that as income increases, the number of people supporting death penalty also increases. A report by Galllup report (Newport 2009) corroborates this finding and shows that households receiving $20,000 to $29,000 are more prone to supporting death penalty but those belonging to the $50,000 to $74,999 have lower propensity towards capital punishment. How strong are the connection between income and support/opposition to death penalty? When we did the a Pearson’s chi-square test, we got Χ2=8.317 and p=.760. This means that the association between the two variables are not statistically significant and that the hypothesis is correct: greater incomes do not assure greater support for death penalty. It is possible that the increase in support for capital punishment among individuals with greater incomes is simply a result of the increasing number of people who are registering higher income levels for the study. Table 3: Crosstabulation, Respondents income and decision to support or oppose capital punishment FAVOR OR OPPOSE DEATH PENALTY FOR MURDER Total FAVOR OPPOSE LT $1000 Count 4 5 9 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 1.4% 3.8% 2.2% $1000 TO 2999 Count 4 2 6 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 1.4% 1.5% 1.4% $3000 TO 3999 Count 4 1 5 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 1.4% .8% 1.2% $4000 TO 4999 Count 4 1 5 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 1.4% .8% 1.2% $5000 TO 5999 Count 3 1 4 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 1.1% .8% 1.0% $6000 TO 6999 Count 5 2 7 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 1.8% 1.5% 1.7% $7000 TO 7999 Count 3 1 4 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 1.1% .8% 1.0% $8000 TO 9999 Count 6 1 7 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 2.1% .8% 1.7% $10000 – 14999 Count 21 16 37 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 7.4% 12.1% 8.9% $15000 – 19999 Count 21 5 26 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 7.4% 3.8% 6.3% $20000 – 24999 Count 35 16 51 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 12.4% 12.1% 12.3% $25000 OR MORE Count 156 73 229 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 55.1% 55.3% 55.2% REFUSED Count 17 8 25 % within FAVOR OR OPPOSE DEATH PENALTY FOR MURDER 6.0% 6.1% 6.0% Reference List Bryman, A., 1988. Quantity and quality in social research New Editio., London: Routledge. Jones, J.M., 2003. Understanding Americans’ support for the death penalty. Gallup. Available at: http://www.gallup.com/poll/8557/understanding-americans-support-death-penalty.aspx [Accessed January 15, 2012]. Newport, F., 2009. In U.S., two-thirds continue to support death penalty. Gallup. Available at: http://www.gallup.com/poll/123638/in-u.s.-two-thirds-continue-support-death-penalty.aspx [Accessed January 15, 2011]. Randall, K., 2000. US death penalty study exposes a system rife with errors. World Socialist Website. Available at: http://www.wsws.org/articles/2000/jun2000/dp-j22.shtml [Accessed January 15, 2012]. Stack, S., 2004. Support for the death penalty: A gender-specific model. Behavioral Science, 43(3-4), p.163-179. The National Data Program for the Sciences, 2011. About GSS. NORC at the University of Chicago. Available at: http://www3.norc.org/GSS+Website/About+GSS/ [Accessed January 15, 2011]. Read More