City-Level Characteristics and Individual Handgun Ownership

• Date: 01 February 2009
• DOI: 10.1177/1043986208329085
• Published date: 01 February 2009
• Subject Matter: Articles

45

Journal of Contemporary

Criminal Justice

Volume25 Number 1

February 2009 45-66

© 2009 Sage Publications

10.1177/1043986208329085

http://ccj.sagepub.com

hosted at

http://online.sagepub.com

City-Level Characteristics and

Individual Handgun Ownership

Effects of Collective Security

and Homicide

Gary Kleck

Florida State University, Tallahassee

Tomislav Kovandzic

University of Texas at Dallas

General Social Survey (GSS) data are used in a multilevel analysis to examine the rela-

tionship between an individual’s decision to own a handgun and his or her city’s

(a) homicide rate and (b) police strength level. The cities in which respondents lived

were identified using special supplementary codes provided by the National Opinion

Research Center so that information about surrounding cities could be attached to each

GSS respondent. Logistic regression analyses indicate that the likelihood of handgun

ownership is increased by higher local homicide rates. The effects are not mediated by

the individual’s own victimization experiences or fear of crime. Positive macro-level

associations previously found between homicide rates and gun ownership levels may

be indicative of homicide effects on handgun acquisition rather than the reverse.Larger

city police forces discourage handgun ownership, supporting the idea that the provision

of greater collective security reduces the felt need of the citizenry to provide their own

protection.

Keywords: gun ownership; fear of crime; victimization; police; homicide rate

America is the most heavily armed nation on earth. It is estimated that by the end

of 2004 there were more than 280 million guns in private hands and that about

half of the nation’s households have a gun (Kleck, 1997, pp. 96-99; Shooting

Industry, 2005; U.S. Bureau of Alcohol, Tobacco and Firearms, 2002). Some have

asserted that this high level of gun ownership is an important cause of the nation’s

high level of violence (Newton & Zimring, 1969; Wiebe, 2003), whereas others have

concluded that although guns affect the character of violence, they have little impact

on its level (Kleck, 1997; Wright, Rossi, & Daly, 1983).

The sources and origins of this extraordinary phenomenon are the subject of con-

siderable dispute. Some contend that American gun ownership is mostly the product

of socialization into a rural hunting subculture having little to do with crime and vio-

lence (Stinchcombe et al., 1980; Wright et al., 1983), whereas others emphasize the

impact of fear, prior criminal victimization, perception of risk, and exposure to high

crime rates (Lizotte, Bordua, & White, 1981; Smith & Uchida, 1988). Still others

46 Journal of Contemporary Criminal Justice

have viewed gun ownership as the pathological product of violent tendencies,

racism, or sexual inadequacy (Sherill, 1973; Stinchcombe et al., 1980;Young, 1985).

For those who see gun ownership as problematic and gun control as an effective

means of violence reduction, it matters a great deal how these disputes are resolved

because they bear on why people own guns and thus how easily they might be

induced to give them up or forego acquiring them. For scholars assessing the impact

of gun ownership rates on violence rates, resolution of these issues is crucial for

determining whether observed associations between the two indicate that guns

cause violence or violence causes gun acquisition. And for scholars seeking to

understand why Americans bring deadly weapons into their homes, these issues are

obviously vital.

When guns are acquired for defensive purposes, this implies that the collective

security promised by the state is perceived as inadequate and that the police in par-

ticular cannot provide sufficient protection from criminal threats. Indeed, some

scholars have conceptualized gun ownership for self-protection as an individual, pri-

vate effort to obtain security and justice, which is to be contrasted with the public

provision of those goods by the state. Thus, a defensive gun is a private substitute

for, or at least a supplement to, public, collectively provided protection in the form

of the police and the rest of the criminal justice system (McDowall & Loftin, 1983).

The traditional argument made by legal and political scholars was that individual cit-

izens give up some of their rights to exact private vengeance and righting of wrongs

in exchange for the benefits of a state collectively providing both justice and protec-

tion. This bargain, however, loses its attractiveness as the effectiveness of the state

in delivering these benefits declines or is perceived as declining.

Thus, people who believe that they cannot rely on the police for security, or who

have little confidence in the crime control effectivenessof police and courts, should

be more likely to own guns. Survey evidence does indeed indicate that people with

less confidence in the ability of the police to provide protection are more likely to

own guns (Feagin, 1970; Lizotte & Bordua, 1980; Smith & Uchida, 1988; Young,

1985). It should, however, be stressed that the effects of confidence in the police

and the effects of actual police strength levels are quite distinct matters, in the

absence of any evidence showing a perfect, or even strong, relationship between

police strength levels and confidence in the police. Although policy makers can

alter the number of police, this may or may not influence levels of public confi-

dence in the police.

These theoretical notions lead to some simple predictions concerning gun owner-

ship. First, gun ownership should be more likely in places with higher crime rates.

This may be because crime leads to fear, which motivates the acquisition of guns,

especially handguns, for protection. Or those personally victimized may respond by

acquiring guns for protection. Alternatively, higher crime rates may motivate acqui-

sition of guns as a prudent precaution, without the effect being meditated by fear or

prior victimization. Second, gun ownership should be less likely in places where

police strength is greater because public confidence in police ability to protect them