FAMILY,ACQUAINTANCE, AND STRANGER HOMICIDE: ALTERNATIVE PROCEDURES FOR RATE CALCULATIONS*

This paper describes the extent of missing data within the Supplementary Homicide Report (SHR), collected as part of the Uniform Crime Reporting Program of the FBI. The yearly SHR provides coded information on the victim, the offender, and the circumstances of all reported homicides in the United States. Thus, the data allow the computation of specific kinds of homicide rates, such as those involving family members, acquaintances, and strangers. However, missing data within reported events, primarily on offender characteristics and thus the victim/offender relationship, present a serious obstacle to the accurate calculation of such rates. The authors propose computational procedures designed to compensate for missing data and empirically evaluate the impact of these procedures on comparative analyses of homicide rates for cities, metropolitan areas, and states.     

Multiple Imputation of the Supplementary Homicide Reports, 1976–2005

The Supplementary Homicide Reports (SHR), assembled by the Federal Bureau of Investigation (FBI), have for many years represented the most valuable source of information on the patterns and trends in murder and non-negligent manslaughter. Despite their widespread use by researchers and policy makers alike, these data are not completely without their limitations, the most important of which involves missing or incomplete incident reports. In this analysis, we develop methods for addressing missing data in the 1976–2005 SHR cumulative file, related to both non-reports (unit missingness) and incomplete reports (item missingness). For incomplete case data (that is, missing characteristics on victims, offenders or incidents), we implement a multiple imputation (MI) approach based on a log-linear model for incomplete multivariate categorical data. Then, to adjust for unit missingness, we adopt a weighting scheme linked to FBI annual estimates of homicide counts by state and National Center for Health Statistics mortality data on decedent characteristics in coroners’ reports for deaths classified as homicide. The result is a fully-imputed SHR database for 1976–2005. This paper examines the effects of MI and case weighting on victim/offender/incident characteristics, including standard errors of parameter estimates resulting from imputation uncertainty.

An Assessment of the Quality of Homicide Data in the Supplementary Homicide Reports: A Research Note

The Federal Bureau of Investigation (FBI)’s Supplementary Homicide Reports (SHR) is a national-level incident-based data clearinghouse for homicide events in the USA used in countless research studies to track homicide trends, test theory, and analyze policy. Despite its wide use, the SHR has well-known limitations. This research note examines the accuracy of the information in the SHR by comparing homicide cases in Newark, New Jersey with their respective data to determine the level of disagreement between the data sources, which variables exhibit the greatest disagreement, and the case-related variables related to the disagreement. Uniquely, we do this for cases where offenders are known in the SHR. Our findings suggest that variables such as victim-offender relationship and circumstance have high disagreement even when the SHR reports an offender, and that the most important covariate of this is time to close the homicide investigation with an arrest. Research implications of the results are discussed.     

WHEN MISSING DATA ARE NOT MISSING: A NEW APPROACH TO EVALUATING SUPPLEMENTAL HOMICIDE REPORT IMPUTATION STRATEGIES*

The Supplemental Homicide Reports (SHR) are widely used in criminological research and inform a broad range of research topics and subsequent policy applications. A serious issue with the SHR is missing information about the offender and incident in many recorded homicides. Although it is convenient to discard cases with missing data before analysis, such discarding is not theoretically justified and can lead to incorrect substantive conclusions. Recently, several techniques for imputing missing SHR data have been proposed, but it is difficult to evaluate their effectiveness. This research presents a new approach to testing and evaluating SHR imputation techniques. Offender data that are missing in the SHR are often found in police records available for individual cities. We examine similarities and differences among cases with known offender characteristics in the SHR, cases with such information missing in the SHR but available in police records, and cases with such information missing in both sources. We then use these data sets to evaluate four different imputation techniques suggested in the literature (Fox, 2004; Messner, Deane, and Beaulieu, 2002; Pampel and Williams, 2000; Regoeczi and Riedel, 2003). We apply each imputation technique to the SHR, and for cases with information missing in the SHR but known in the police records, we see how well the imputed values correspond both with the individual known values and with the overall distributions in the police records. We discuss the outcome of our assessment of these strategies, and we outline important implications this assessment has for research using SHR data.     

CIRCUMSTANCES IN SUPPLEMENTARY HOMICIDE REPORTS: VARIETY AND VALIDITY*

This paper presents a critical examination of homicide circumstances as reported in supplementary homicide reports (SHR). Different types of homicides can be distinguished by the circumstance codes and victim/offender relationship recorded on the SHR. Delineating murder by type invites analysis of this offense from a victimization perspective–homicides have much in common with nonlethal offenses. Different types of homicide present different policy problems to police. The paper also discusses various sources of error in SHR data, which must be recognized by researchers interested in theoretical or policy questions.     

Economic Development,Change of Age Distribution,and Stream Analogy of Homicide and Suicide: A Cross-National Assessment

The current study tests the stream analogy of lethal violence, proposed by Unnithan et al. on a large sample of 124 nations. Spatial econometric regression models confirm that economic development decreases homicide rate and increases suicide–homicide ratio (SHR). The SHR developed by Unnithan et al. accounts for the impact of economic development better than original measure of suicide rate. The most important finding of the current investigation is that the increase in the population size of the elderly, consequent to economic development, is partially responsible for elevated suicide rates and SHRs.     

Male and female juveniles arrested for murder: a comprehensive analysis of U.S. data by offender gender

Murders committed by juveniles remain a serious concern in the United States. Most studies on juvenile homicide offenders (JHOs) have used small samples and have concentrated on male offenders. As a result, little is known about female JHOs and how they differ from their male counterparts on a national level. This study utilized the Supplementary Homicide Report (SHR) database to examine more than 40,000 murders committed by male and female juvenile offenders from 1976 to 2005. This research effort, the most expansive to date, replicated previous findings with respect to gender differences using bivariate and multivariate analyses. As predicted, six variables used to test eight hypotheses with respect to male and female JHOs in single-victim incidents were significant (victim age, victim-offender relationship, murder weapon, offender count, victim gender, and homicide circumstance). Regression analysis revealed that all variables remained significant when entered into the model. This article concludes with a discussion of our findings and directions for future research.     

HOMICIDE AGAINST THE ELDERLY: A RESEARCH NOTE*

By virtue of the type of data generally used (victim surveys), previous research on the victimization of the elderly is limited in two respects. Not only is the crime of homicide outside the domain of victimization data, but sample surveys uncover too few incidents of victimization of the elderly to permit in-depth analyses. Using the supplementary homicide reports from 1976–1985, we compared patterns and rates of homicide among the elderly and younger populations. Our results suggest that the elderly are in fact the least at risk for homicide generally, as has been reported for other crimes. When examining specific subtypes of homicide, however, the elderly are actually at greater risk than their younger counterparts for homicide committed during a robbery.     

The Application of Missing Data Estimation Models to the Problem of Unknown Victim/Offender Relationships in Homicide Cases

Homicide cases suffer from substantial levels of missing data, a problem largely ignored by criminological researchers. The present research seeks to address this problem by imputing values for unknown victim/offender relationships using the EM algorithm. The analysis is carried out first using homicide data from the Los Angeles Police Department (1994-1998), and then compared with imputations using homicide data for Chicago (1991-1995), using a variety of predictor variables to assess the extent to which they influence the assignment of cases to the various relationship categories. The findings indicate that, contrary to popular belief, many of the unknown cases likely involve intimate partners, other family, and friends/acquaintances. However, they disproportionately involve strangers. Yet even after imputations, stranger homicides do not increase more than approximately 5%. The paper addresses the issue of whether data on victim/offender relationships can be considered missing at random (MAR), and the im-plications of the current findings for both existing and future research on homicide.     

The social ecology of homicide: A modified lifestyle/routine activities perspective

Recent research on criminal victimization suggests that lifestyle may offer important insights into the unequal distribution of crime. The lifestyle approach is used here to explore variations in homicide rates for Dade County, Florida. The application of this approach to the analysis of cross-sectional variations in homicide rates requires (1) elaboration of the assumptions that connect the likelihood of victimization for individuals to variations in the rate of victimization by location as well as (2) justification of the selection of specific variables as indicators of variation in exposure to the risk of homicide. In sum, four individual characteristics and six residential characteristics are said to be implicated in the explanation of variations in homicide rates among census tracts. Census tract data derived from 1980 Bureau of Census files are used to explore these variations. Findings suggest that future researchers should consider the relative importance of individual attributes, especially marital status, in increasing the risk of homicide victimization.     

Criminal recidivism in sexual homicide perpetrators

Forensic psychiatric reports on 166 sexual homicide perpetrators in Germany were retrospectively analyzed for criminal risk factors. Follow-up information about release and reconvictions from federal criminal records was available for 139 offenders; 90 (64.7%) had been released. The estimated recidivism rate (Kaplan-Meier analyses) for 20 years at risk was 23.1% for sexual and 18.3% for nonsexual violent reoffences. Three men (3.3%) were reconvicted for attempted or completed homicide. Only young age at the time of sexual homicide resulted in higher sexual recidivism, whereas increased nonsexual violent recidivism was related to previous sexual and nonsexual delinquency, psychopathic symptoms, and higher scores in risk assessment instruments. Increased recidivism with any violent reoffence was associated with age-related factors: young age at first sexual offence, at homicide, and at release and duration of detention. The impacts of the results for risk assessment, relapse prevention, and supervision are discussed.     

Comparing Homicide‐Suicides in the United States and Sweden

Research on homicides followed by suicides has largely relied on very localized samples and relatively short time spans of data. As a result, little is known about the extent to which patterns within cases of homicide‐suicides are geographically specific. The current study seeks to help fill this gap by comparing twenty years of homicide‐suicide data for Sweden and a large U.S. county. Although some of the underlying patterns in the two countries are similar (e.g., decreasing rates), a number of important differences emerge, particularly with respect to incidence, weapons used, perpetrator age, and relationship of the perpetrator to the victim.     

THE STRUCTURAL COVARIATES OF URBAN HOMICIDE: REASSESSING THE IMPACT OF INCOME INEQUALITY AND POVERTY IN THE POST-REAGAN ERA*

Numerous studies have explored the relationship between rates of homicide and income inequality and poverty. However, a general consensus on the theoretical and empirical connections among these variables has yet to be reached. This article reports the findings of a city-level analysis of this relationship, using 1990 data for the 190 largest cities in the United States. In order to address several methodological and theoretical concerns in prior literature, three separate measures of inequality and three categories of disaggregated homicide rates are analyzed. The results suggest that both inequality and poverty have significant and independent positive effects on rates of homicide in U.S. cities following the largest increase in the economic gap between rich and poor in our nation's history.     

Structural correlates of female homicide: A cross-national analysis

The present study went beyond previous cross-national homicide research, which has largely focused on combined (male and female) rates of homicide offending, by using gender-disaggregated homicide arrest figures. The study included controls for the clearance rate and the percentage of homicides that were attempts, and included data for forty-eight countries across multiple levels of development. The author compared the effects of development/modernization and opportunity on female homicide rates to their effects on male homicide rates. Results indicated that, overall, structural predictors had very similar effects on male and female homicide rates. Both rates were lower in countries with a higher gross domestic product (GDP) per capita, and neither male nor female homicide rates were related to urbanization. Countries with a higher number of people per household had a lower rate of both male and female homicide offending; however, this relationship only held when percent young was excluded from the model.     

On urban homicide: A statistical analysis

A statistical analysis is made of homicide rates in the 50 largest American cities for four different years. It is shown that differences in recent murder growth among the cities can largely be explained as typical random fluctuations about a common trend. It is also found that the changing age profile of the American people explains no more than ten percent of the increase in homicide since 1964. Several mathematical models for future homicide growth are proposed from the analysis, and under each the probability of death by murder and corresponding drop in life expectancy are estimated for individuals born now in each of the 50 cities.     

STRUCTURAL COVARIATES OF U.S. COUNTY HOMICIDE RATES: INCORPORATING SPATIAL EFFECTS*

Spatial analysis is statistically and substantively important for macrolevel criminological inquiry. Using county‐level data for the decennial years in the 1960 to 1990 time period, we reexamine the impact of conventional structural covariates on homicide rates and explicitly model spatial effects. Important findings are: (1) homicide is strongly clustered in space; (2) this clustering cannot be completely explained by common measures of the structural similarity of neighboring counties; (3) noteworthy regional differences are observed in the effects of structural covariates on homicide rates; and (4) evidence consistent with a diffusion process for homicide is observed in the South throughout the 1960–1990 period.     

Social support and homicide in transitional Russia

Like the United States, Russia is a large industrialized nation with high violence rates. Although its overall homicide rate is among the highest in the world, however, local rates of crime vary widely. Similarly, the level of social support provided by the state varies throughout Russia due to former Soviet policies, the differential pace of political and economic change, and the level of development. Relying upon recent criminological literature on social support theory, this study tested the hypotheses that areas with higher levels of social support will have lower homicide rates and that the effects of negative socioeconomic change on homicide rates will be moderated by levels of social support. Utilizing data from Russian regions (n = 78) and controlling for other structural covariates, negative binomial regression was employed to estimate the effects of social support on regional homicide rates. As expected, negative socioeconomic change was associated with higher homicide rates, but the results provided no support for direct or conditioning effects of social support on homicide. The findings are discussed in the context of Russia-specific conditions and of the meaning of these findings for recent research on social support and crime.     

Spousal Homicide in Russia Versus the United States: Preliminary Findings and Implications

Increased research on spousal homicide warrants a cross-cultural comparison that Russia and the United States could provide. As a first step, official statistics and scholarly reports are summarized in terms of spousal homicide rates, sex-ratios of spousal homicide, and circumstances of these homicides and inclusion of attempted homicides in Russia. The statistics are adjusted to compensate for several methodological limitations in establishing homicide data, such as misclassifications of homicides. Adjusted homicide data suggest that Russia has a higher spousal homicide rate, more female victims, and fewer shootings than the United States. Women in Russia may be two and one-half times more likely to be killed by their spouses or lovers than their counterparts in the United States. The break-up of the Soviet Union and contradictory status of women in Russia may contribute to these findings. Future research might include homicide case reviews and perpetrator assessments to substantiate and refine these preliminary findings.