Gun Ownership and Attitudes Toward Gun Control in Older Adults: Re-examining Self Interest Theory
Department of Sociology, Anthropology, and Criminal Justice Studies, Longwood University, Farmville, Virginia, United States of America
Previous research that examines attitudes towards gun control consistently argues that self-interest is an important predicting factor in favoring or opposing gun control measures. In previous research, this has been understood and tested in terms of owning a gun resulting in opposition to gun control (Wolpert and Gimple 1998; O’Brien et al 2013). However, this relationship has not been specifically tested by age. While research shows a positive correlation between age and a desire to own guns, no research directly tests self-interest within older adult populations who own guns. The current research addresses the question; does self-interest predict attitudes towards gun control in older adult populations who currently own guns? This research uses the 2012 wave of the General Social Survey to test self-interest and attitudes toward gun control specifically in older adult populations. Those in the 70-79 cohort are the most likely to report owning a gun (45.5%), while also having one of the highest rates of support for gun control (76.6%), disproving what has been suggested about self-interest and gun control. A binary logistic regression model shows this relationship remains even after controlling for other factors, like political ideology.
Gun Control, Gun Ownership, Attitudes, Self-Interest
Received: August 4, 2015
Accepted: August 16, 2015
Published online: September 17, 2015
@ 2015 The Authors. Published by American Institute of Science. This Open Access article is under the CC BY-NC license. http://creativecommons.org/licenses/by-nc/4.0/
At a time in the United States when new conversations about gun control appear almost daily, it becomes paramount to understand how potential voters feel about gun control measures in the United States. After all, gun control has clearly been on President Obama’s agenda. On December 14th 2012, after the Newtown school shooting, where 26 people including 20 young students were killed in gun violence, President Obama remarked, "If there is even one step we can take to save another child, or another parent, or another town, from the grief that has visited Tucson, and Aurora, and Oak Creek, and Newtown, and communities from Columbine to Blacksburg before that – then surely we have an obligation to try." By January 16th, 2013 President Obama with the help of Vice President Biden delivered policy proposals that included measures of gun control (whitehouse. gov, "Now is the time to do something about gun violence."). The authors find that age has an important, yet ignored, role in attitudes about gun control policy in the United States. This is especially significant for those interested in older adult populations because of their high levels of voter turnout (Bhatti, Hansen, Wass 2012).
Gun control is a highly contested issue in the United States. This is not a new issue, however. Over the past 30 years, research has addressed attitudes toward gun control in numerous ways. Studies have examined attitudes by political ideology (Wolpert and Gimpel 1998; Celinska 2007), race and racial attitudes (O’Brien, Forrest, Lynott, Daly 2013), gender (Celinska 2007), religious beliefs (Celinska 2007), and region of residence (Brennan, Lizotte, and McDowall 1993; Burger 2002). However, researchers have not looked at the impact of age on attitudes toward gun control. Elderly persons actively participate in voting and other activities that help in influencing matters of the state and country, and have been credited in the media for influencing policy change on many social policies, including health care and social care services. Researchers have argued the strength of elderly influence within the United States in terms of resolving specific issues through political means is greater than any other age group within the United States (Bhatti, Hansen, Wass 2012). Therefore, understanding how this age group feels about gun control is important for political and social reasons.
Several researchers have argued that those who own guns are more likely to be opposed to gun control measured. In other words, these researchers use a self-interest framework as an explanation, sometimes the primary explanation, for opposing gun control. Based on this theoretical understanding of what predicts attitudes toward gun control and the lack of research that addresses how age influences attitudes toward gun control in the United States, the current research addresses the following question: how does age impact how people feel about gun control? Findings indicate that the self-interest argument does not explain attitudes toward gun control attitudes in older cohorts of individuals. Specifically, those over the age of 50 are more likely to support gun control measures versus any other age group, even when they report owning at least one gun.
2. Literature Review
Previous research suggests that self-interest strongly influences individuals view on gun control (Kleck 2009; O’Brien et al. 2013; Swift 2013; Wolpert & Gimple 1998). According to Kleck (2009), in 1990, more than 200 million guns could be found inside the borders of the United States and most of the owners were against gun control measures. Research often uses gun ownership to operationalize the influence of self-interest on gun control attitudes. This is intuitive because owning a gun is a personal choice based on the attitudes and interest of the person who owns the gun.
Self-interest has been dominant in the view of psychology and in much of Western thought. For example, Thomas Hobbes believed that human beings always acted from self-interest. More specifically, Hobbes believed that everything we do ultimately was for our own benefit (Andre and Velasques 2013). Researchers have long posited that self-interest affects opinion on government policies. For example, in sociological and economic theory, self-interest was discussed by Marx (1887/1927) in terms of materialist theory and then debated by Weber (1921/1968) in terms of instrumental rationality (Breznau 2010; Author 2013). This approach asserts that attitudes originate from individual’s calculations on what policies and practices will have the most personal benefit to them. This benefit can come in the form of financial or personal benefits (Andreb and Heien 2001). As such, people are less likely to favor policies or practices that hurt them financially or where they feel their economic loss does not outweigh the social gains (Brooks and Manza 2007). This theory has garnered both support and criticism in contemporary literature.
Self-interest can be operationalized as owning a gun because gun ownership is a choice that is made when individuals weigh the consequences of having a potentially lethal weapon in their home versus the perceived benefit of having that weapon. According to Glaeser and Glendon (1998), the people who are most likely to own guns are either criminals, hunters, or people who own guns for self-protection. People who live in high crime areas tend to see gun ownership as a necessity for their protection. Specifically, In 2013, sixty percent of gun owners reported that their primary use for obtaining a firearm was for personal protection (Swift 2013). In lower crime areas, people report purchasing guns when they have felt personally threatened by physical violence (Glaeser and Glandon 1998). Glaeser and Glandon’s (1998) findings also indicate that white men who have higher incomes, and have teenage children have the highest probability of owning a gun especially if they also report a distrust with the local law enforcement. Some research has gone as far as to suggest that increasing gun availability can reduce crime because potential victims deter potential criminals (Altheimer 2012). In a related study, Lott (2013) finds that many gun owners question how/if the implementation of gun control policies would be effective in eliminating accidental shootings, suicides, and homicides. Therefore, previous research suggests that the motivation in obtaining a firearm is often associated with self-protection or hunting, where individuals believe the benefits of owning a gun based on their self-interest outweighs the potential consequences.
Citizens across the United States have diverse reasons as to why they support or oppose gun control. However, self-interest has continually been mentioned in research as one of the most likely factors in people’s attitudes (Kleck 2009; O’Brien et al. 2013; Swift 2013; Wolpert & Gimple 1998). What has been ignored by previous research, however, is the impact of age on how self-interest influences attitudes toward gun control.
Age and Gun Control
Very little research examines the relationship between gun control attitudes and age. Elderly people, however, have been shown to be more likely to own guns than younger people. According to Kleck (2009), the desire to possess a firearm for personal protection increases as individuals advance in age. More research, however, addresses the use and ownership of guns by elderly people. Statistics have shown that people 65 years and older have higher odds of having a gun in their home than any other age group. A 2007 study stated that 27% of people ages sixty-three and older had a firearm in their home (Hepburn, Miller, Azrael, and Hemenway 2007). Burbick (2006) describes this phenomenon as a result of aging white men who grew up on Western and civil war notions of life starting to see themselves as these American heroes and purchase more weapons to fit that role.
Data show a multitude of elderly persons apply to carry a concealed weapon every year, and as of 2010 these numbers were expected to continually rise. Interestingly, because of this steady increase of elderly persons purchasing guns, the firearms industry decided to create a gun specifically for elderly people. The gun known as the Palm Pistol is a single-chamber firing mechanism that could be fired via squeeze ball instead of the standard trigger. The gun is specifically designed for persons with limited manual dexterity and seen to be perfect for senior citizens. The firearms industry even tried to register the gun as a medical device so that Medicare would pay for it, but the Food and Drug Administration rejected the proposal. With the creation of an easier operational hand weapon, elderly persons that own a firearm are expected to continually increase (Mertens and Sorenson 2012).
The reason for elderly persons wanting to own a firearm has not been well established by the literature. A New York Times article suggests several reasons why elderly might be more likely to own guns. First, the majority older persons felt they needed a firearm for self-defense. Additionally, guns, like many other possessions, are seen as a sign of independence and individualism to many Americans. However, this reason is disputed by some because often adult children of aged adults who own firearms have increased concern for their parents’ safety because they fear an increased risk of accidental shootings or suicide. Lastly, guns seem to have strong sentimental meaning to many older adults, making it difficult to nearly impossible to remove the gun from their possession (Russakoff 2010).
According to aggregate data from the General Social Survey, 69.5% of Americans between the age of 31 and 40, favor gun permits compared to 80.2% of Americans between the ages of 56 and 89. One possible explanation of this is presented by Dowler (2002) who finds that people who get their news from the television are far more likely to have a stronger Anti-gun control attitudes. Those Americans who do not watch the television for their source of news, and receive their information from the newspaper, tend to be more in favor of gun control. One way that Dowler (2002) accounts for this difference is that news printed in the paper tends to have more detailed information and clearly depicts how the crime happened. Other research addresses a similar issue with information available online. Neves, Amaro, and Fonseca (2013) found that only 10% of a sample of people over the age of 64 in Lisbon reported using the internet. Thus, one possible explanation for differences in attitudes about gun control is that elderly are more likely to read newspapers compared to younger people who may rely more heavily on internet and television news sources.
In 2013, Fiorina, Abrams, & Pope reported that 64% of Democrats favor stricter gun control and 52% of Republicans also favored gun control. Gun control and ownership is a topic that has been prevalent in politics within the American society for quite some time. According to Celinska (2007) the individual that is most likely to own a firearm would be a Republican male. Likewise the individual that is least likely to own a firearm Democrat female. A similar study conducted by Carter (2002), used the 2000 wave of the General Social Survey (GSS) and came to the conclusion that 55% of individuals that identified as conservatives reported owning either a revolver or another type of firearm in their home while only 22% of the individuals that identified as liberal reported owning a firearm in their home. This information was contrary to what was found by Lott in 1988. According to Lott, the ownership of a firearm was proportional amongst the two political super powers (Lott 1988). In summary, most current empirical research with the exception of Lott (1988), suggests that Democrats are more likely than Republicans to support stricter gun control while Republicans are more likely to own guns.
Previous research suggests gender and sex differences in regards to attitudes toward gun control. Izadi (2013) reports that 65% of women favor stronger gun laws, compared to 44% of men. In addition, Smith (1984) suggests that 35% of men and 46% of women believe that pistols and revolvers should be banned and only accessed by police officers. Additionally, women are far more likely than men to say that mass shootings could be reduced if there were a ban on assault weapons (Bell 2013). Not only do woman favor gun control more than men in the United States, but according to Celinska (2007) women in the United States favor gun control more than woman in the United Kingdom due to their influence towards violence and their inability to deter crime. Other factors that may contribute to women having more favorable attitudes toward gun control measures are that women report being more concerned with the environmental and social risks associated with guns. This includes things like accidents and being overpowered by others who have guns (Kahan & Braman 2003).
Another way that gender and gun ownership has been studied is through qualitative interviews with white men. In these interviews, Stroud (2012) finds that in order to fulfill their masculine roles men admit to carrying weapons. Specifically, Stroud (2012) finds that many men feel that carrying a weapon in public embodies what it means to be a man, a good husband, and a protective father. As these fathers and husbands age, the possession of a weapon allows them to still feel protective, regardless of increasing physical limitations due to the aging process. These older men, who were above forty years of age at the time of the interviews, said that they felt vulnerable when not allowed to carry a gun.
Some research has found a relationship between socioeconomic status (SES) and attitudes toward gun control. Research that addresses SES finds that people with higher educational attainment and income are more likely to support gun control legislation (Wright 1981). Lott (1988) disagrees in part with these findings by saying that the poor and less educated are most likely to support gun control. However, in regards to actually attaining permits, Kleck (1996) finds that the support for gun permits is strongest amongst individuals whom on average have a higher-income, and more schooling than their counter-parts. Constanza (2004) agrees with Kleck by stating concealed carry permits are more commonly supported in wealthier homes and higher economic areas. The wealthier are more likely to acquire a concealed carry permit, whereas the less wealthy are less likely to acquire one. This relationship may influence attitudes toward other types of gun control (Constanza, 2004).
Burbick (2006) discusses the ideals of a "Gun Show Nation" in regards to how and why America has become obsessed with guns and gun control. According to Burbick (2006) the "Gun Show Nation" is based on the historical idea of white, older, southern men needing to protect their daughters from criminals and black predators. Burbick (2006) uses many famously white and masculine men, like Teddy Roosevelt, Buffalo Bill, and Arnold Schwarzenegger to show how masculinized and racialized the perceptions of guns have become. Further, according to Celinska (2007) on average the individual that is most likely to own a firearm would be a white, married, middle-aged male. Likewise the individual that is least likely to own a firearm would be an African American, middle aged, female (Celinska, 2007).
Few research studies have been conducted examining attitudes toward gun control and race. Recently, however, a group of researchers studied how racist ideologies may influence attitudes towards gun control and gun ownership. O’Brien et al. (2013) argued that those in the United States who had higher levels of racist ideologies are more likely to be opposed to gun control and more likely to own guns. This research was conduced using the 2006 wave of the National Election Survey and did not control for regional variations. The composite data from the years of 1972-2012 in the General Social Survey shows that 75.2% of white respondents favor gun permits, compared to 82.4% of African Americans who favor gun permits, and 85.5% of those of all other races favoring gun permits.
Southerners have been found to have higher levels of gun ownership than other regions nationwide. As a result, Southerners are also the most opposed to gun permits, handgun bans, and gun registration. This attitude is directly seen through cultural differences (Brennan, Lizotte, & McDowall 1993; Dolak 2013). Several studies also argue that people in the west are also likely to own guns and oppose gun control (Stroud, 2012).
Gun laws also vary by state. Specifically, there are 13 states and the District of Columbia that require a police permit in order to purchase any sort of gun. The states that require the purchaser of the weapon to have a permit are: Connecticut, District of Columbia, Hawaii, Illinois, Iowa, Massachusetts, Michigan, Minnesota, Mississippi, Nebraska, New Jersey, New York, North Carolina, and Ohio (NRA, 2014).
The current study examines age differences in attitudes toward gun control laws in the United States. Data comes from the 2012 wave of the General Social Survey (GSS). Gun control attitudes is measured by a question that asks respondents if they believe a permit should be acquired from the police before a person can purchase a firearm. The GSS also asks respondents if they have a gun in their home. This is the question we use to measure self-interest, using the blueprint of previous research. The 2012 wave of the GSS includes 1940 respondents in a nationally representative sample. We omitted cases with missing data on the gun control/ownership variables. We handle missing data on explanatory variables by including a missing data dummy variable. Therefore, our final sample size on the individual level is 1240 respondents.
The initial analysis is a cross tabulation that examines attitudes toward gun control by age and gun ownership. For this cross tabulation, age is coded into cohorts of 10 years, with the exception of 18 and 19 year olds, who are included in the 20-29 year old cohort. After results were examined for the initial cross tabulation, the sample was limited to only those who reported owning guns and then the second cross tabulation was created. These two initial cross tabulations are Figure 1 and Figure 2.
After examining the descriptive statistics, logistic regression models were run to control for the other variables that previous research shows impacts attitudes toward gun control. This was done to ensure that the relationship between age and gun control attitudes was consistent even after other impacting variables were controlled for. Again, the first iteration of the regression included the full sample and included gun ownership as a dummy variable, and then non-gun owners were eliminated from the sample and another regression model was run. Model 1 in Table 1 is based on individual level variables. Model 2 in Table 1 includes all of the variables from Model 1 and also includes regional level variables as level two variables. Table 2 consists of the full model only (the same as Model 2 in Table 1 but with non-gun owners eliminated from the sample). All coding for variables is described.
Measures of Independent Variables – Level 1
The primary focus of this research is how age influences attitudes toward gun control specifically testing self-interest when a respondent owns a gun. In the cross tabulation, age was grouped into six categories: 18-29, 30-39, 40-49, 50-59, 60-69, 70-79, and 80 and above. The regression models include age as a continuous variable with one-year increments to allow for more specific information on how attitudes shift as age increases gradually.
Other independent variables were included that are commonly used to understand attitudes toward gun control and gun ownership, as discussed in the background section of this research. In this study, these variables are used to control for other influences on attitudes and ownership.
Race was measured in three categories, white (reference group), Black, and Other race. The vast majority of GSS respondents were white, which limited the variation available for analysis. To compensate for this limitation, weights were applied to adjust for more accurate generalization.
Gender has also been shown to influence attitudes toward gun control (Izadi 2013). Therefore the measure "sex" (female = 1, male = 0) is included in the regression analysis. Because sex is such a strong predictor of attitudes toward gun control gender role ideology variables were also examined to see if they influenced attitudes toward gun control. Because no significant relationship was found on any level, and because previous research does not mention this relationship, these variables were not included in the final models, but analysis is available upon request.
Lastly, two variables were used to measure political ideology theory. Political ideology is measured both in terms of political party identification and degree of anti-statist sentiment. In the United States, as noted above, the Republican Party is more conservative and anti-statist than the Democratic Party (Lipset, 1997). In addition, many citizens choose not to affiliate with either political party. To measure political party identification, four dummy variables were created: Republican, which was used as the reference group, Democrat, independent, and "other." Respondents who answered "other" could have either left or right leanings, as was true for independents. To measure political ideology in terms of strength of anti-statist, a variable generated from a question that asked respondents if the United States needs more government, less government, or neither more nor less government was included in the models.
Several demographic controls were also included. Social class was measured in two ways. The first was a self-report measure of subjective social class (working, middle, or upper) and the second was education. Education was coded: high school degree or less, some college education, and a college degree or higher.
Lastly, a variable that asked respondents if they thought the United States was spending too much on Black people, specifically was included. This is similar to a measure of racial attitudes used by O’Brien et al. (2013).
Level 2 variables
Clustering is used in the regression models to better understand the fix and random effects of attitudes at the regional level. The GSS has a region variable that includes several subgroups. This variable has been recoded this into 6 regions: the Northeast, the Mid-Atlantic, the South, the Midwest, and the West.
Because we are interested in how age predicts gun control attitudes, we use a multilevel logistic regression modeling strategy where data is clustered at the regional level. Therefore, the individual-level variables are entered at level 1 and the regional-level variable is entered at level 2. Clustering is a basic type of generalized linear mixed models, also known as multilevel or hierarchical models. These models are "mixed" because they allow fixed and random effects, and they are "generalized" because they are appropriate for continuous Gaussian responses as well as binary, count, and other types of limited dependent variables (Cohen 2003). It was established that a multilevel model must be used in this research because the intra-class correlation between regions is significantly different. In other words, the between country variation for the dependent variables is significantly different from one another and thus without multilevel modeling some variation would be ignored. Therefore, a technique called clustering is used to group (or cluster) countries in the analysis, thus, addressing the nonindependence of within-state observations (Tai and Treas 2009).
In Figure 1, percentages of those who own guns are shown in the first column (black). In the youngest age group 23.4 percent of people report owing a gun, this is the lowest percent of any age group. The highest percent of people who own guns is represented by the 70-79 year olds (45.5 percent). The next column (light gray) represents those who support gun control in the form of requiring a police permit to purchase and own a gun. The lowest support for this measure is in the 30-39 year old population where 66 percent of people support gun control measures. The highest level of support occurs in the 80 and over age group where 81.4 percent of people support control (36.1 percent of this population reports owning a gun). This figure shows that as age increases so does support for gun control. A general trend occurs whereas age increases so does the rate of gun ownership with the exception of those 80 and over.
Data: 2012 wave of the General Social Survey
Figure 2 shows attitudes toward gun control for only those who stated they own a gun. Again, as age increases so does the support for more gun control even when only examining the attitudes of those who own guns. Specifically, only 50.1 percent of 20-29 year olds who own guns support gun control where as 66.0 and 66.7 percent of 70-79 and 80 and over year olds support gun control, respectively. The highest level of support for gun control by gun owners occurs in the 60-69 year olds where 67.1 percent of gun owners support gun control. It should also be noted that 67.0 percent of 50-59 year olds support gun control. As in the previous figure, there is a general trend that as people age, even when they own guns, they become more in favor of gun control.
Data: 2012 wave of the General Social Survey
Table 1 shows the results of logistic regression models. Previous research demonstrates that there are multiple variables that impact attitudes toward gun control. Therefore, several regression models were run to ensure that the relationship between age and gun control remains even after controlling for other impacting variables. Overall, findings indicate that there is a significant relationship between age and attitudes toward gun control after controlling for other variables. Specifically, the odds of supporting gun control increase by 1.7 for every year increase in age. Overall, people who own guns have 43.2 percent lower odds of supporting gun control. This is significant (p<.001) and consistent with previous research. Females have 79 percent higher odds of supporting gun control compared to men. Democrats have 89 percent higher odds of supporting gun control compared to Republicans. Lastly, respondents who feel the government should have more influence, compared to less influence, have higher odds of supporting gun control.
|Model 1||Model 2|
|Education (College Degree)|
|Political Ideology (Republican)|
|Gov Influence (Less gov)|
Notes: Data comes from the 2012 wave of the GSS
Significance: ***p<.001; **p<.005; *p<.05
When regional level variables are included, findings indicate people in New England, the Mid-Atlantic and the Mid-West have significantly higher odds of supporting gun control compared to those in the South. There was no difference between respondents in the South compared to those in the west. This is similar to the findings of Blendon, Young, and Hemenway (1996) who also found similarities between the South and the West when it comes to gun control opinions.
|Education (College Degree)|
|Political Ideology (Republican)|
|Gov Influence (Less gov)|
Notes: Data comes from the 2012 wave of the GSS
Significance: ***p<.001; **p<.005; *p<.05
Overall, this resembles the findings of previous research with one exception. In O’Brien et al. (2013) they found racist attitudes was a predictor of attitudes toward gun control, however, the variable in the GSS used to measure subjective racism was not a significant predictor of gun control attitudes. However, this investigation is not the primary purpose of the current research.
In Table 2 the same variables that were used in Model 2 of Table 1 were included in analysis. However, non-gun owners were eliminated from the sample. This left 432 respondents. Again, this was done to ensure that the relationship found between age and gun control attitudes remained when only gun owners were examined. Findings indicate a significant positive relationship between age and attitudes toward gun control (OR=1.015, p<.05). Again, this relationship was in a model that controlled for other variables that can impact attitudes toward gun control.
While researchers have been interested in attitudes for gun control for decades, few have made any real attempt at understanding how age impacts these attitudes. The current research shows that as age increases so does the likelihood that a person supports gun control in the form of requiring a police permit to own a gun. This is true for gun owners and non-gun owners alike. While researchers have previously found support for self-interest theory when applied to gun control, these findings would suggest that there is an age component to this theory. Self-interest theory does not explain why people over the age of 50 support gun control. This argument can be made based on the high levels of people over the age of 50 who both own guns and support gun control.
Using clustered logistic regression models we find that the relationship between age and attitudes toward gun control remains even when controlling for other variables that impact attitudes toward gun control. Along with age, gender, political ideology, and region of residence are significant predictors of attitudes toward gun control, this is true for a full sample that represents residents in the United States and a sample that examines only gun owners.
Given the current political conversation about gun control in the United States it is important to understand how the populous feels about gun control. This research suggests that most people in the United States would be in favor of requiring police permits to own a gun. This is especially true in the population over the age of 50, who also have high levels of owning guns and of voting participation. By 2050, about one-fifth of the total United States population will be sixty-five years of age and older. By 2030 most of the increase that will be taking place will slow down, when the last of the baby-boomers reach the age of sixty-five (Jacobsen, Kent, Lee, and Mather 2011). The population aging has many implication for the United States, these implications has been discussed by aging researchers, but the findings of the current research suggest that this could potentially have implications for gun control in the United States.
Using the GSS, the only type of gun control attitude that is measured is requiring police permits, and, admittedly, this is one of the lesser forms of gun control that has been suggested by politicians. However, it is a type of gun control that is not required currently in most of the United States, which makes it an interesting starting point for research. Future research should examine why self-interest is not adequate to explain attitudes toward gun control in older populations and the impact of stronger forms of gun control.