Attitudes toward capital punishment: Educational, demographic, and neighborhood crime influences

Abstract

Studying attitudes toward capital punishment has been a topic of interest for decades. Indeed, it is often the subject of Gallup polls, political commentary, and social science research. Research indicates that attitudes vary by demographic factors, educational influences, and neighborhood crime rates. Building on prior research, the current study examined death penalty attitudes among 599 college students and 213 residents from high crime and low crime neighborhoods. In particular, the research investigated differences between how students and residents view the death penalty as no research to date has analyzed this issue. The authors also analyzed how different populations perceived the impact of race on the administration of the death penalty. Results suggested that demographics, especially race, have the strongest impact on death penalty attitudes, but that there are few differences between students and residents and those living in high and low crime neighborhoods.

Keywords:

• Capital punishment
• Death penalty
• Public opinion
• College students
• Race

Notes

1 Tel.: [phone]; fax: [phone].

2 Tel.: [phone].

3 We defined “high crime” neighborhoods as those which had crime rates at least two standard deviations above the mean, and “low crime” neighborhoods as those which had crime rates at least 2 standard deviations below the mean.

4 We utilized a three step sampling strategy in order to identify potential respondents for the survey. First, neighborhoods were initially identified using census block groups. Next, streets were randomly selected within those census block groups. Finally, households were randomly chosen from the streets identified in step two, providing the final sample in which surveys were distributed (Browning, 2002).

5 Due to the low response rate for residents, we compared respondents from the ten “high crime” neighborhoods and the ten “low crime” neighborhoods to census data on key demographic variables. In terms of age, gender, and divorce rate, survey residents are quite comparable to the census (with females being slightly overrepresented, as is common in survey research). In terms of differences, we found that whites were overrepresented among survey respondents in high crime neighborhoods, though they were approximately equivalent in low crime neighborhoods.

6 In fact, newer research has found that lower participation in surveys has not dramatically reduced the validity of the results (Keeter, Miller, Kohut, Groves, & Presser, 2000; Curtin, Presser, & Singer, 2000). Indeed, at least one recent study has shown that attempts to increase participation actually increased survey error (Merkle & Edelman, 2009).

7 We only used six of the ten questions. The standard death penalty question asking about using the death penalty for convicted murderers was omitted so that results from the present study could be compared to results obtained from Coogle's research as part of a broader research project that is not described in the current article.

8 Collinearity diagnostics (e.g., Variance Inflation Factors) were analyzed to see whether multicollinearity was a problem and results indicate that there are no problems with multicollinearity.

9 Due to a high correlation between marital status and race among residents, there was reason to believe that an interaction was occurring; thus an interaction term was computed and entered into the model (race × single).