ArticlePDF Available

Structural Covariates of Homicide Rates: Are There Any Invariances Across Time and Social Space?



This study demonstrate that the empirical literature on the structural convariates of homicide rates contains inconsistent findings across different time periods and different geographical units. This apparent variance of findings may be due to statistical or methodological artifacts of particular studies, such as different time periods covered, units of analysis, samples, model specification, and problems of statistical analysis and inference. A baseline regression model using 11 structural covariates is estimated for cities, metropolitan areas, and states in 1960, 1970, and 1980. The empirical estimates of this model exhibit instability because of high levels of collinearity among several regressors. Principal components analysis is applied to simplify the dimensionally of the structural covariate space. Reestimation of the regression model then indicates that the apparent inconsistencies across time and social space are greatly reduced. The theoretical significance of the findings for substantive theor...
Structural Covariates of Homicide Rates: Are There Any Invariances Across Time and Social
Author(s): Kenneth C. Land, Patricia L. McCall, Lawrence E. Cohen
The American Journal of Sociology,
Vol. 95, No. 4, (Jan., 1990), pp. 922-963
Published by: The University of Chicago Press
Stable URL:
Accessed: 21/05/2008 22:36
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We enable the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact
... Some scholars have elaborated upon IAT to explicate the social-psychological characteristics of anomie. As Konty (2005) argues, the cultural elements of anomie result in "microanomie," which is a value system that promotes self-interest over the collective well-being. Similarly, Groβ et al. (2018) have developed the concept of "market mentality" to operationalize IAT at the individual level. ...
... Specific to the current analysis, microanomie/market mentality are expected to affect the antisocial behavior of adolescents. Konty (2005) finds that the self-enhancing values associated with microanomie contribute to a greater degree of acceptance of deviant behavior as well as more cheating, vandalism, larceny, assault, alcohol intoxication, and marijuana abuse amongst adolescents. Additionally, Groβ et al. (2018) find that market mentality corresponds with higher rates of selfreported delinquent behavior. ...
... The main difference between these models is at the country level. Because there are more issues associated with multicollinearity at higher levels of aggregation (Land et al., 1990), we examine different model configurations to confirm that our results are robust. While the bivariate correlation between only two of the variables warrants moderate concern about multicollinearity (r = .68 ...
The present study examines cross-national variation in school-based bullying victimization. Specifically, we address whether decommodification, a concept implicated in Institutional Anomie Theory that measures the degree of a society’s social welfare protection, is a protective factor against school-based bullying victimization. To test this theory, we retrieve data from the Programme for International Student Assessment (PISA) questionnaire and combine this data with other sources capturing cross-national factors hypothesized to impact bullying victimization. The sample consists of 286,871 adolescents (with an average age of 15 years) attending 14,192 schools nested within 55 high-and-middle-income countries. We estimate multilevel regression models with three levels of analysis (student, school, and country), finding that countries with a greater degree of decommodification have lower rates of school-based bullying. Overall, our findings illustrate that the national level of social welfare protection, which had been previously neglected in this research literature, is a robust predictor of bullying victimization.
... Contrary to popular opinion and common sense, property offending is not related to income level, either at the individual or aggregate level (Elliott et al., 1985;Krivo & Peterson, 1996;Rekker et al., 2015; though see Jarjoura et al., 2002 demonstrating the pervasive effects of persistent childhood poverty on all kinds of youth offending). On the other hand, violent offending is closely linked with poverty, both at the individual level (Cunradi et al., 2000;Mok et al., 2018;Rekker et al., 2015) and various aggregate levels (Kaylen et al., 2017;Land et al., 1990;McCall et al., 2010;Messner et al., 2001;Parker & Pruitt, 2000;Tcherni, 2011). ...
Full-text available
The increasing rates at which psychotropic drugs have been prescribed to children and adolescents in the USA in the last three decades (since the early 1990s) have prompted questions about whether this trend is associated with the “great American crime decline.” Medicalization can be considered one of the strategies to remedy children’s neuropsychological deficits and improve their self-control. Another possible remedy is school-based services for children with learning disabilities, mandated by the Individuals with Disabilities Education Improvement Act (IDEA) of 2004. Using state-level panel data analyses for years 1990–2014 (with the main focus on 2000–2014 outcomes), the current study estimates associations between these two developmental self-control remedies—medicalization and school-based services—and minor, moderate, and severe types of juvenile violence, while controlling for relevant covariates (both time-varying and time-invariant). The results of mixed-effects linear regression analyses accounting for powerful time trends show a strong association between increases in school-based services for children with learning disabilities and declines in all types of juvenile violence. Another strong and consistent finding that emerges in the analyses is the link between reductions in child poverty at the state level and decreases in juvenile violence, both contemporaneously and over time. Psychotropic drug prescribing to children (measured using Supplemental Security Income rolls of children with mental health conditions) exhibits inconsistent or insignificant effects. The findings of this study have substantial theoretical and policy implications and indicate the importance of strengthening school-based services for children with disabilities and reducing child poverty as essential violence prevention tools.
... Structural theories can thus assist us to understand male violence in South Africa. Such theories require that we examine what are referred to as strain and control (see Land et al., 1990;Pridemore, 2002;Sampson and Lauritsen, 1994). While strain theories concentrate on how structural inequality drives people to commit violent crime, control theories posit that violent crime is committed when social bonds are weak. ...
This chapter reviews theories on the immigration–crime link, some macro-level, some micro-level, including those that pose a positive relationship between immigration and crime as well as those that pose a negative relationship between immigration and crime.
A plea for scaling up (i.e., examining the global dimensions) and expanding out (i.e., considering indirect, alternative and complementary measures) of local crime and criminal justice phenomena is made in this paper. The implications and benefits of doing so are illustrated by utilizing the study of unresolved (or cold case) homicides, which are usually conceived as events responsive to localized co-victim, police, and prosecutorial actions. Cross-nationally, the homicide rate is found to be significantly correlated to the formal contact rate and conviction rate (i.e., as indicators of criminal justice system responsiveness to such cases). Inter-woven with the discussion is a personal narrative of how the author came to develop an interest in this and related topics.
Full-text available
Background: Pakistan is a country where the rate of crime is constantly increasing, with many ex-convicts committing the crime again after getting released from prison, with no knowledge of the antecedents of their recidivism or their psychological consequences. The current study was conducted in order to develop a theory about the social and environmental causes of recidivism along with the psychological consequences for Pakistani recidivist ex-convicts. Methods: The sampling technique used was theoretical sampling, amounting to a total sample of 20 recidivist ex-convicts before reaching the saturation point. A grounded theory method was incorporated for the current research as the objective of the study was the development of a theory on the social and environmental causes of recidivism along with its psychological consequences. A semi-structured interview protocol was used for the purpose of data collection. Whereas the interview transcripts were coded in three phases, including open, axial, and selective coding for theory generation. Results: The individuals shared their detailed experiences of social and environmental factors which contributed to recidivism, and three main antecedents were identified, including negative social experiences, hatred for society, and a toxic environment. Additionally, several psychological symptoms as a consequence of the negative social experiences were also found. Conclusions: The findings of the study have added new information to the concept of recidivism as there were no previous theoretical concepts pertaining to recidivism in the context of Pakistan. These identified antecedents can be useful in strategizing interventions in Pakistan, as currently there are no such intervention programmes in Pakistan due to a lack of prior studies on the subject matter.
Structural disadvantage has long been empirically linked to violent crime across different race/ethnic groups. More recently conceptualized as “racial invariance,” observed racial differences in crime rates are hypothesized to be the result of disparities in community-level structural conditions. However, most investigations into this hypothesis have focused on urban settings, with limited attention to rural contexts. The current study seeks to fill this gap by comparing county-level structural predictors of homicide victimization for Black, White, and Latinx populations in both urban and rural communities. Consistent with the racial invariance hypothesis, findings reveal that disadvantage strongly predicts homicide across race/ethnicity in both rural and urban counties. Closer inspection of results, however, exposes noteworthy differences in the effects in rural and urban settings.
Full-text available
Cohen and Felson's (1979) theory of “routine activities” is evaluated using 1972 and 1980 crime rate data for 676 American suburbs. Cross-sectional and change influences of criminal motivation (percent poor, percent unemployed, percent black, and percent youth), criminal opportunities (employment concentration and percent multiple housing) and guardianship (police employment, police expenditure, and female labor force nonparticipation) on suburban crime rates, 1980 and change in crime rates, 1972–80 are estimated using maximum likelihood (LISREL) techniques. Consistent with the routine activities model, a multiplicative effect of the preconditions for crime on crime rates is also estimated. Motivation, opportunities, and guardianship are found to have direct and/or indirect additive effects on violent and/or property crime rates—supporting the “routine activities” model. A multiplicative effect of the preconditions of crime is found for property crime but not for violent crime rates.
Full-text available
Criminal opportunity theory provides a framework for examining the interaction between potential offenders and potential victims. Criminals' behavior influences the nature and amount of self-protective measures taken by potential victims, and changes in self-protection make criminal opportunities more or less attractive. Criminal opportunity theory has precursors in criminological theory, preeminently in the work of Cloward and Ohlin, but in these theories opportunities are mediated through social learning. Criminal opportunity theory employs the economic theory of markets to describe and predict how criminals and victims interact. Evidence is available that potential victims take more self-protection measures when the perceived risk of victimization is greater and that prospective criminals are likelier to attack relatively more vulnerable targets. Little research is available on whether increases in self-protection reduce the total volume of crime or merely displace crime to more vulnerable targets; the extent of displacement probably differs among offenses. The market perspective has several benefits to the investigation of interaction between potential victims and offenders: it assembles different topics encompassed by criminal opportunity theory into a coherent whole, it is expressed in a form that facilitates borrowing from economic theory, and it generates new and important insights for policy evaluation and criminological theory. One central insight is that law enforcement strategies may alter the quality of opportunities and thereby precipitate additional crime. Effective incapacitation or rehabilitation policies, for example, may reduce the number of offenders in circulation and thereby reduce the perceived risk of victimization. This may cause individuals to reduce their self-protection efforts, making them more attractive targets than before and thereby stimulating increased crime rates on the part of those criminals who remain active.
This study systematically tests a formal theory of how certain dimensions of social stratification--income, race, and age--relate to risk of predatory criminal victimization. An opportunity theory of criminal victimization is proposed, focusing on the mediating role played by five risk factors: exposure, guardianship, proximity to potential offenders, attractiveness of potential targets, and definitional properties of specific crimes themselves. Propositions are derived pertaining to the bivariate and multivariate-partial (main) effects expected from the theory and tested in analyses based on a representative sample of the U.S. population for the crimes of assault, burglary, and personal larceny. These data indicate that the relationship between the dimensions of social stratification and the offenses studied here is complex, and that, other things being equal, those usually thought to be most vulnerable economically and socially--the poor, the nonwhite, the old--are not the most likely victims of crime. Race has little direct effect on victimization risk, while age is inversely related to each type of crime at both the bivariate and multivariate levels of analyses. The findings are largely consistent with the proposed theory.
Studies have consistently shown the South to have the highest homicide rates in the United States. Two theories to explain this have been proposed: a regionally based subculture of violence thesis and a structural interpretation emphasizing high rates of poverty in the South. These explanations are not mutually exclusive, but their proponents have engaged in a sometimes rancorous debate. Investigations designed to test the two theories have produced inconsistent findings. The present study attempts to resolve some of the questions by using OLS and ridge regression techniques to analyze state homicide rates for total populations, whites and nonwhites. Our findings show homicides for total populations and whites to be influenced by both poverty and regional differences. Among nonwhites, however, poverty is not related to the homicide rate, and the presumed effect of southern culture depends on the measurement adopted.. The results lend support to arguments that the high homicide rates of white southerners and blacks reflect, in part, subcultural differences.
This paper asks whether decreases in the proportion of the U.S. population in high crime-prone ages in recent years have produced decreases in crime rates which correspond approximately, but in an inverse direction, to the increases observed when the baby boomers reached the high crime-prone ages in the 1960s and 1970s. We examine age distributions of arrests for murder and motor vehicle theft and then specify and estimate structural-equation models of the time trajectories in annual rates of these crimes for the post-World War II period. With these models, we test the hypothesis that the relationship of crime to age composition is symmetric and infer that it cannot be rejected. We also find that simple models that contain effects of trends in the age structure, business cycles, trends in criminal opportunity, and the rate of imprisonment can account for most of the variance in annual rates of homicide and motor vehicle theft from 1946 through 1984. We discuss some implications of our findings for forecasting U.S. crime rates for the remainder of this century and for theories of the macrodynamics of crime causation.
The growing research on comparative studies of homicide in the United States reveals significant methodological advances but inconsistent findings. A major goal is to identify sources of inconsistency and accumulate more valid and reliable results. This analysis empirically examines a major problem with most previous comparative studies--the failure to disaggregate the overall homicide rate into more refined and conceptually meaningful categories of homicide. A theoretically integrated model is presented that guides the calculation of disaggregated rates and the derivation of hypothesized relationships. Using data from the Comparative Homicide File (CHF), the analysis shows that indicators of resource deprivation and social disintegration tend to have significant effects across subtypes of homicide, although the magnitude of the effects varies, while indicators of violent cultural orientation are confined to homicides resulting from interpersonal conflicts. The implications of the results for comparative research on homicide are discussed.
The most prominent explanation for the disproportionate involvement by blacks in criminal violence is the subculture of violence theory (Wolfgang and Ferracuti, 1967; Curtis, 1975). Recently, several studies have attempted to test the subculture of violence thesis by examining the direct effect of racial composition on the aggregate homicide rate in multivariate analysis. However, the aggregate homicide rate fails to distinguish the effects of racial composition from environmental context, thereby confounding the interpretation of community and individual-level influences. These studies thus have been unable to test key sociological propositions derived from subcultural theory. In this article, we address the race and crime issue by examining the structural sources of variation in age-, race-, and sex-specific rates of homicide offending in major U. S. cities.
This article presents statistical materials bearing on the ques tion of whether the variation in city crime may be resolved into a relatively small number of statistical components which have a recognizable sociological meaning. Two basic series of data are included: (1) seven major offense rates and 28 social, econ nomic and demographic characteristics for U.S. cities, 100,000 and over, 1950; (2) seven major offense rates and 26 social, eco nomic, and demographic characteristics for U.S. cities, 100,000 and over, 1960. The intercorrelations among variables in each series are given, along with their respective loadings after rota tion on the first five factors as obtained by the method of prin cipal components. Partial correlations with age, sex, and num ber of inhabitants held constant and their principal factors are also given. The results of the factor analysis disclosed that the incidence of crimes against the person was related to a group factor which was tentatively designated "minority relations," and that the incidence of crimes against property was related to a group fac tor tentatively named "anomie." Taking both minority relations and anomie as indices of normative conflict and confusion, the general conclusion is reached that city crime varies with the degree of differential social disorganization.