Gun Culture: Mapping a Peculiar Preference for Firearms in the Commission of Suicide

Abstract

Research has shown how honor cultures promote aggression against others (e.g., homicide) and the self (e.g., suicide). Two studies examine the connection between honor and a predilection for guns in the commission of suicide. Study 1 shows that Whites living in honor states are especially likely to use guns to commit suicide, controlling for gun accessibility. Study 2 reveals that a “gun access gap” in honor states—a positive difference between the proportion of all suicides that involve a gun and the gun ownership rate—predicts outcomes linked previously to honor cultures: homicides and accidental deaths.

Notes

¹As noted by Nitobe (1899/2002), the ancient Japanese ritual of seppuku involved a Samurai stabbing himself in the stomach (which was believed to be the seat of the soul) prior to being beheaded. By not showing his pain to the audience, the Samurai could demonstrate the purity of his soul and thus restore his stained honor.

²Some past studies on culture of honor have included White Hispanics with White Anglos, whereas others have not. We did not include Hispanics in the present study, in part because this demographic group is so unequally distributed across the United States that we did not wish to confound this variable with our categorization of honor and nonhonor states. However, we should note that analyses that included Hispanics with White Anglos did not differ appreciably from those reported here.

³The question regarding firearm ownership was not asked in California in 2002. Thus, we used only responses from the 2001 survey for California.

⁴All data obtained from CDC were calculated per 100,000 persons and adjusted for age based on the U.S. standard population in the year 2000.

⁵When non-White firearm suicide rates for Alaska, New Hampshire, and Vermont are treated simply as missing data, the interaction is still significant despite the reduction in power, F(1, 41) = 27.22, p < .001.

Note. R ² Whites = .81; R ² non-Whites = .69. CH = culture-of-honor status (0 =non-honor state, 1 =honor state); Economic Deprivation = economic deprivation composite index; HPSA = proportion of population living in a health professional shortage area; Temperature = winsorized mean annual temperature.

*p ≤ .05. **p ≤ .01.

⁶With New Hampshire, Vermont, and Alaska's data points missing, states’ CH status was still not a significant predictor of the non-White firearm suicide rate, β = .001, t(41) = .01, p = .99. Also, the firearm suicide rate for non-Whites was still skewed after winsorizing Alaska because of a few other outliers. We log-transformed the rate, which eliminated the positive skew, and used this new rate as the outcome variable in a separate regression analysis. The result was nearly identical to the one we report in the text with the merely winsorized rate; CH status was not a significant predictor of the non-White firearm suicide rate (β = .05, p = .63), but the non-White gun ownership rate was, β = .80, t(44) = 7.34, p < .001.

⁷Following Cohen (1998), argument-related homicide cases included lovers’ triangles, brawls under alcohol, brawls under drugs, arguments over money, and other arguments. Although Cohen (1998) specified an age range of 15 to 39 for his analyses, we were unable to obtain population estimates that perfectly matched this age range. We thus substituted the available age range of 15 to 34 for both the offender data and the population estimates.

⁸Analyzing argument-related homicide rates with the gun access gap that excludes Hispanic Whites did not change any of our conclusions.

⁹Before the log transformation, the argument-related homicide rates were skewed at 1.41. After the transformation, skew was more acceptable at 1.07.

Note. All variables are untransformed. Gap (Whites/non-Whites/Whites including Hispanics) = gun access gap for Whites/non-Whites/Whites including Hispanics; Homicide (Whites/non-Whites) = age-adjusted homicide rates per 100,000 for Whites/non-Whites; Accidents (Whites/non-Whites) = age-adjusted deaths associated with unintended, external causes per 100,000 for Whites/non-Whites; Economic Deprivation = economic deprivation composite index; HPSA = proportion of population living in a health professional shortage area; Temperature = winsorized mean annual temperature; Rurality = proportion of population living in rural areas.

*p ≤ .05. **p ≤ .01.

¹⁰Because the non-White accidental death rate was quite skewed at 1.78, we performed a natural log transformation on this covariate prior to analysis. The resulting skew was a more acceptable 0.60.

Note. Whites: R ² Homicides =.76; R ² Accidents =.75; R ² Argument homicides =.59. Non-Whites: R ² Homicides = .44; R ² Accidents =.63. When the outcome is White total homicides, or accidents, the gun access gap for Whites is used, and when the outcome variable is non-White homicides or accidents, the gun access gap for non-Whites is used. When the outcome is White argument homicides, the gun access gap for Whites that includes those of Hispanic origin is used. HPSA = proportion of population living in a health professional shortage area; Temperature = winsorized mean annual temperature; Rurality = proportion of population living in rural areas. For White total homicides and accidents, non-White homicides and accidents were used as covariates, respectively. For non-White homicides and accidents, White homicides and accidents were used as covariates, respectively.

*p ≤ .05. **p ≤ .01.