Search

The United States initiated sweeping counterterrorism efforts after the September 11, 2001 terrorist attacks. This study tests a backlash hypothesis as it relates to the nature of attacks against the US abroad. Relying on data from the Global Terrorism Database, this study uses a quasi-experimental design to investigate whether attacks against the US abroad became more or less lethal after 9/11. There is a significant increase in the proportion of attacks with fatalities and a significant decrease in the proportions of non-lethal attacks against US targets and interests overseas after 9/11. The results suggest a redistribution in the lethality of attacks against the US abroad. This study finds evidence of a backlash of deadlier terrorism violence against the US abroad after September 11. Examining for unintended consequences is an important facet of terrorism prevention research and policy.

The U.S. homicide rate increases of 2015-2016 occurred within a longer recorded history of trend variations. This article examines the place of the 2 years within this history and considers if they were unusual. An analysis of more than five decades of panel data finds that 2015 and 2016 were, in fact, somewhat unusual, but not far outside the normal range of variation. The analysis also provides no evidence that changes like those in 2015-2016 can predict future trends. As a more general matter, this article argues for increased attention to the typical characteristics of crime rate changes.

This paper discusses the perception that displacement in terrorism is inevitable; that antiterrorism efforts merely relocate terrorism in some way. Using quarterly time-series data from the Global Terrorism Database (1994–2013) and the vector autoregression framework, we test the following hypothesis: the target-hardening efforts within the United States (US) after September, 11, 2001 reduce attacks on domestic US targets, but increase attacks on US targets abroad. To provide a more comprehensive test, we also provide dynamic impact factors and variance decompositions. The results of this intervention analysis show no support for displacement, and instead provide support for a diffusion-of benefits hypothesis. We also discuss how criminological and especially crime-prevention knowledge can guide and encompass the study of terrorism.

Objectives:This study examines whether the use of target-hardening measures engenders greater amounts of casualty terrorist attacks against protected targets. Specifically, this study evaluates the impact of augmenting aviation security and protection of U.S. embassies and diplomats on the frequency and proportion of casualty attacks against aviation targets and U.S. diplomatic targets, respectively.Method:Using time-series data from the Global Terrorism Database (1970 to 2001), this study conducts time-series intervention analysis. To provide a more comprehensive test, a variety of supplementary analyses—consisting of data transformations, various onsets of the interventions, autoregressive integrated moving average, Poisson, and vector autoregression models of time-series data—are performed.Results:We found no increase in the frequency or proportion of casualty attacks against protected targets following target-hardening interventions. The results show that the typical ensuing terrorist attack against hardened targets is not violence based (i.e., maximizing casualties).Conclusions:Findings that attacks against hardened targets did not become deadlier provide support for the criminological message that unintended harmful effects from situational terrorism prevention strategies are the exception rather than the rule.

Chapter 6 introduces two conceptual issues that, in our opinion, will become important in the near future. The first involves the validity of statistical inference. Critics of the conventional null hypothesis significance test generally focus on the undue influence of sample size on p-values and the common misinterpretation of significance levels. Bayesian approaches address and, to some extent, solve both shortcomings. The second conceptual issue involves the use of control time series. As a rule, valid causal inferences require the use of a contrasting control time series. In most instances, no ideal control series is available; however, a synthetic ideal control series can sometimes be constructed from an ensemble of less-than-ideal control time series.

Chapter 4 introduces the full ARIMA intervention model. Most substantive theories specify the intervention as an exogenous dichotomy. A Box-Tiao transfer function then distributes the intervention's response across the endogenous time series to reflect a theoretically specified onset and duration. Transfer functions allow the noise component to be parsed from the residualized time series. Theoretical specification of the intervention model requires at least some sense of the onset and duration of the impact. Detailed analyses of ten time series demonstrate how to handle interventions with abrupt and permanent, gradually accruing, gradually decaying, and complex impacts. One popular version of an ITSA short course ends with Chapter 4. Although statistically adequate ARIMA models can be built using the modeling strategy described in Chapters 3-4, survey knowledge of the auxiliary methods described in Chapter 5 is recommended.

Interrupted Time Series Analysis develops a comprehensive set of models and methods for drawing causal inferences from time series. Example analyses of social, behavioural, and biomedical time series illustrate a general strategy for building AutoRegressive Integrated Moving Average (ARIMA) impact models. The classic Box-Jenkins-Tiao model-building strategy is supplemented with recent auxiliary tests for transformation, differencing and model selection. New developments, including Bayesian hypothesis testing and synthetic control group designs are described and their prospects for widespread adoption are discussed. Example analyses make optimal use of graphical illustrations. Mathematical methods used in the example analyses are explicated assuming only exposure to an introductory statistics course. Design and Analysis of Time Series Experiments (DATSE) and other appropriate authorities are cited for formal proofs. Forty completed example analyses are used to demonstrate the implications of model properties. The example analyses are suitable for use as problem sets for classrooms, workshops, and short-courses.

Chapter 5 describes three sets of auxiliary methods that have emerged as add-on supplements to the traditional ARIMA model-building strategy. First, Bayesian information criteria (BIC) can be used to inform incremental modeling decisions. BICs are also the basis for the Bayesian hypothesis tests introduced in Chapter 6. Second, unit root tests can be used to inform differencing decisions. Used appropriately, unit root tests guard against over-differencing. Finally, co-integration and error correction models have become a popular way of representing the behavior of two time series that follow a shared path. We use the principle of co-integration to define the ideal control time series. Put simply, a time series and its ideal counterfactual control time series are co-integrated up the time of the intervention. At that point, if the two time series diverge, the magnitude of their divergence is taken as the causal effect of the intervention.

Chapter 2 introduces ARIMA algebra. With a few exceptions, this material mirrors the authors’ earlier work. The chapter begins with stationary time series processes – white noise, moving average (MA), and autoregressive (AR) processes – and moves predictably to non-stationary and multiplicative (seasonal) models. Stationarity implies that the time series process operated identically in the past as it does in the present and that it will continue to operate identically in the future. Without stationarity, the properties of the time series would vary with the time frame and no inferences about the underlying process would be possible. A seasonally nonstationary process drifts or trends in annual steps. The “best” seasonal model structure is the one that transforms the series to white noise with the fewest number of parameters.

Chapter 3 develops the methods or strategies for building ARIMA noise models. At one level, the iterative identify-estimate-diagnose modeling strategy proposed by Box and Jenkins has changed little. At another level, the collective experience of time series experimenters leads to several modifications of the strategy. For the most part, these changes are aimed at solving practical problems. Compared to the 1970s, for example, modelers today pay more attention to transformations and to the usefulness and interpretability of an ARIMA model. The Box-Jenkins ARIMA noise modeling strategy is illustrated with detailed analyses of twelve time series. The example analyses include non-Normal time series, stationary white noise, autoregressive and moving average time series, nonstationary time series, and seasonal time series. The time series models build in Chapter 3 are re-introduced in later chapters.

Chapter 1 introduces Interrupted Time Series Analysis (ITSA) as a toolbox for researchers whose data consist of a long sequence of observations | say, N ≥15 observations | measured before and after a treatment or intervention. Sometimes the treatment or intervention is implemented by the researcher, other times it occurs naturally or by accident. The chapter also describes a family of impact types, characterized by their onset (abrupt or gradual) and duration (permanent or temporary); and the essential role of counterfactual controls in causal inference. The chapter concludes with an outline and summary of the book's subsequent chapters.

To evaluate the National Violent Death Reporting System (NVDRS) as a surveillance system for fatal shootings of civilians by law enforcement in the United States.We cross-linked individual-level mortality data from the 2015 NVDRS and 5 open-source data sets ( FatalEncounters.org , Mapping Police Violence, the Guardian's "The Counted," Gun Violence Archive, and The Washington Post's "Fatal Force Database"). Using the comprehensive cross-linked data set, we assessed the proportion of study-identified fatal police shootings that were captured by NVDRS, overall and by state, and by each open-source data set.There were 404 unique study-identified fatal shootings by law enforcement in the 27 states for which data were available from NVDRS, 393 (97%) of which were captured in NVDRS. The proportion of shootings captured by NVDRS varied only slightly by state.The NVDRS provides a comprehensive count of fatal police shootings. Public Health Implications. Expanding NVDRS to all 50 states would provide comprehensive counts of fatal police shootings and detailed circumstantial information about these deaths at the national level. Open-source data can continue to provide real-time data collection as well as more complete information about nonfirearm officer-involved deaths.

Pinker makes the sweeping argument that violence declines with modernization as countries of the world gradually converge in terms of economic markets, communication structures, and culture. An alternative prediction is provided by conflict perspectives, which argue that growing income inequality both within and between countries will serve to drive criminal violence ever higher. We use fixed effects regression models to examine the extent to which national homicide victimization rates for 55 countries have shared declining trends for the past six decades. Our results show that homicide rates in the countries examined trended in the same downward direction since the early 1990s, but support for a modernization perspective was limited mostly to a subset of wealthy, western-style democracies.

This study assesses whether recent victimization influences one’s confidence in the criminal justice system. Specifically, the study tests whether the predictors of confidence in major types of actors in the justice system are different for victims and nonvictims. British Crime Survey (BCS) data are analyzed using a series of structural equation models with multiple group methods. Overall, the findings support the idea that there is little difference between victims and nonvictims in how views of legal actors predict overall confidence. Additionally, prosecutors are slightly more influential on system confidence, regardless of victimization experience. Policy implications and future research directions are discussed.

No policy designed to prevent firearm violence is more popular than mandatory sentence enhancements for gun crimes. By providing stiff and certain penalties when a gun is involved in an offense, sentence enhancement laws' promise to reduce the use of firearms by criminals. Because the laws apply only when a crime is committed, they impose no direct costs on legitimate gun owners. Opinion polls find that a large majority of the public favors mandatory sentence enhancements, and more than half the states have adopted them.2 If these laws deliver their expected crime preventive effects, they are an especially attractive approach to regulating the use of firearms.

The goal of Chapter 2 is to derive the properties of common processes and, based on these properties, to develop a general scheme for classifying processes. Stationary processes includes white noise, moving average (MA), and autoregressive (AR) processes. MA and AR models can approximate mixed ARMA models. A lag or backshift operator is used to solve ARIMA models for time series observations or random shocks. Covariance functions are derived for each of the common processes.Maximum likelihood estimates are introduced for the purposes of estimating autoregressive and moving average parameters.

Chapter 7 begins with an outline and description of five threats to internal validity common to time series designs: history, maturation, instrumentation, regression, and selection. Given the fundamental role of prediction in the modern scientific method, scientific hypotheses are necessarily causal. After an outline of the evolving definition of “causality” in the social sciences, contemporary Rubin causality or counterfactual causality is introduced. Under the assumption that subjects were randomly assigned to the treatment and control groups, Rubin’s causal model allows one to estimate the unobserved causal parameter from observed data. Control time series are chosen so as to render plausible threats to internal validity implausible. An appropriate control time series may not exist, however, an ideal time series may be possible to construct. Synthetic control group models construct a control time series that optimally recreates the treated unit’s preintervention trend using a combination of untreated donor pool units.

Chapter 1 introduces three categories of time series designs: descriptive, correlational, and interrupted time series designs. The evolution from a two-validity system to a four-validity system (including internal, external, statistical conclusion, and construct validities) is then described. Situations where the added expense of a time series design is warranted fall into two overlapping categories. The first category consists of situations where the nature of the underlying phenomenon is obscured by trends and cycles. A well-constructed time series model may reveal the nature of the underlying phenomenon. The second category consists of situations where a known intervention and an appropriate control time series are available. In those situations, a well-designed time series experiment can support explicitly causal inferences that are not supported by less expensive before-after designs. The chapter concludes with an outline and summary of the book’s subsequent chapters.

Chapter 8 focuses on threats to construct validity arising from the left-hand side time series and the right-hand side intervention model. Construct validity is limited to questions of whether an observed effect can be generalized to alternative cause and effect measures. The “talking out” self-injurious behavior time series, shown in Chapter 5, are examples of primary data. Researchers often have no choice but to use secondary data that were collected by third parties for purposes unrelated to any hypothesis test. Even in those less-than-ideal instances, however, an optimal time series can be constructed by limiting the time frame and otherwise paying attention to regime changes. Threats to construct validity that arise from the right-hand side intervention model, such as fuzzy or unclear onset and responses, are controlled by paying close attention to the underlying theory. Even a minimal theory should specify the onset and duration of an impact.

A threat to external validity is any factor that limits the generalizability of an observed result. Unlike all threats to statistical conclusion and internal validities and some threats to construct validity, threats to external validity cannot ordinarily be controlled by design. Nor is there any disagreement on how threats to external validity should be controlled. In most instances, it can only be controlled by replication?—across subjects, situations and time frames. This seldom happens, unfortunately, because the academic incentive structure discourages replication. The contemporary “reproducibility crisis” was spurred by a collaborative group of social scientists attempting to replicate one hundred experimental and correlational studies published in three mainstream psychology journals. Sixty percent of replications failed to reproduce the published effect. Failures to control for threats to external validity that stem from uncontrolled variations in persons, situations, and time frames, parsimosniously explain the failure rate in this replication study.

The general AutoRegressive Integrated Moving Average (ARIMA) model can be written as the sum of noise and exogenous components. If an exogenous impact is trivially small, the noise component can be identified with the conventional modeling strategy. If the impact is nontrivial or unknown, the sample AutoCorrelation Function (ACF) will be distorted in unknown ways. Although this problem can be solved most simply when the outcome of interest time series is long and well-behaved, these time series are unfortunately uncommon. The preferred alternative requires that the structure of the intervention is known, allowing the noise function to be identified from the residualized time series. Although few substantive theories specify the “true” structure of the intervention, most specify the dichotomous onset and duration of an impact. Chapter 5 describes this strategy for building an ARIMA intervention model and demonstrates its application to example interventions with abrupt and permanent, gradually accruing, gradually decaying, and complex impacts.

Chapter 4 downplays forecasting’s role in the design and analysis of time series experiments and emphasizes its potential abuses. While the “best” ARIMA model will outperform other forecasting models in the short and medium-run, long-horizon ARIMA forecasts grow increasingly inaccurate with diminished utility to the forecaster. Although the principles of forecasting help provide deeper insight into the nature of ARIMA models and modeling, the forecasts themselves are ordinarily of limited practical value. Forecasting can provide useful guidance to analysts choosing between two competing univariate models. While forecasting accuracy is only one of many criteria that might be considered, other things being equal, it is fair to say that a statistically adequate model of a process should provide reasonable forecasts of the future. Forecast accuracy depends on a host of factors, many of which lie outside the grasp of model adequacy. More important, forecast accuracy has no universally accepted metric.

Chapter 6 addresses the sub-category of internal validity defined by Shadish et al., as statistical conclusion validity, or “validity of inferences about the correlation (covariance) between treatment and outcome.” The common threats to statistical conclusion validity can arise, or become plausible through either model misspecification or through hypothesis testing. The risk of a serious model misspecification is inversely proportional to the length of the time series, for example, and so is the risk of mistating the Type I and Type II error rates. Threats to statistical conclusion validity arise from the classical and modern hybrid significance testing structures, the serious threats that weigh heavily in p-value tests are shown to be undefined in Beyesian tests. While the particularly vexing threats raised by modern null hypothesis testing are resolved through the elimination of the modern null hypothesis test, threats to statistical conclusion validity would inevitably persist and new threats would arise.

Chapter 3 introduces the Box-Jenkins AutoRegressive Integrated Moving Average (ARIMA) noise modeling strategy. The strategy begins with a test of the Normality assumption using a Kolomogov-Smirnov (KS) statistic. Non-Normal time series are transformed with a Box-Cox procedure is applied. A tentative ARIMA noise model is then identified from a sample AutoCorrelation function (ACF). If the sample ACF identifies a nonstationary model, the time series is differenced. Integer orders p and q of the underlying autoregressive and moving average structures are then identified from the ACF and partial autocorrelation function (PACF). Parameters of the tentative ARIMA noise model are estimated with maximum likelihood methods. If the estimates lie within the stationary-invertible bounds and are statistically significant, the residuals of the tentative model are diagnosed to determine whether the model’s residuals are not different than white noise. If the tentative model’s residuals satisfy this assumption, the statistically adequate model is accepted. Otherwise, the identification-estimation-diagnosis ARIMA noise model-building strategy continues iteratively until it yields a statistically adequate model. The Box-Jenkins ARIMA noise modeling strategy is illustrated with detailed analyses of twelve time series. The example analyses include non-Normal time series, stationary white noise, autoregressive and moving average time series, nonstationary time series, and seasonal time series. The time series models built in Chapter 3 are re-introduced in later chapters. Chapter 3 concludes with a discussion and demonstration of auxiliary modeling procedures that are not part of the Box-Jenkins strategy. These auxiliary procedures include the use of information criteria to compare models, unit root tests of stationarity, and co-integration.

Today the international community seems at a loss as to how to transact peace between Israel and Palestine (and Syria). UN Security Council Resolution 242 of 1967 provides the principles for that peace. Yet there has always been a perceived ambiguity about its withdrawal clause. Diplomatic and UN records show clearly what the Security Council intended in Resolution 242. Nine of 15 members wanted total withdrawal, and the minority saw the virtue of small adjustments to the 1949 Armistice Line to accommodate Israel's demand for �secure and recognized� borders. Every Security Council member upheld the overarching principle, �the inadmissibility of the acquisition of territory by force.� Those who drafted Resolution 242 seem not to have checked that its terms were consonant with the Fourth Geneva Convention, even though they recognized the Convention applied. The Convention renders it illegal for those under occupation to agree terms with the Occupying Power which infringe the rights and protections of the Convention. Since the Convention remains in force until the end of occupation, no peace agreement which includes the adjustment of borders or ceding territory may be concluded until after a full withdrawal has taken place�a requirement fully consonant with Resolution 242's �inadmissibility� principle, and removing any doubt regarding the requirement for a full Israeli withdrawal. To comply with it themselves and to avoid misapprehension, Quartet members must tell Israel, Syria and Palestine that they cannot recognize a peace agreement which would violate the Convention's terms.

We investigated the precision of Supplementary Homicide Report (SHR) homicide rate estimates for large U.S. cities during the 5 years, 1998-2002. The homicide rates based on the National Vital Statistics System provided a parallel measure and the basis for estimating the reliability and average error. When cities with incomplete SHR data were removed from the sample, the estimated reliability was high (.99), and the standard error of measurement was low (1.2 homicides per 100,000 residents). Reliability remained high for subsets of cities and for most subsets of victims. For some groups, however, such as African Americans and persons age 0 to 14, the reliability was much lower.

Criminal victimization is known to influence households’ moving decisions, but theories suggest that the processes leading to a moving decision can vary across racial and ethnic groups. Drawing from current literature, we hypothesized that victimization would have a stronger effect on moving decisions for Whites than for Blacks or Hispanics, and that racial/ethnic residential segregation would moderate the impact of victimization on mobility. Using a longitudinal sample of 34,134 housing units compiled from the National Crime Victimization Survey for the 40 largest metropolitan areas in the United States (1995–2003), we found results that both support and contradict the hypotheses. Specifically, White residents display consistent evidence that victimization is a significant predictor of household mobility. Blacks and Hispanics, in contrast, are more varied in their moving behavior after victimization. In addition, significant differences exist among these groups in responses to victimization and in how mobility is influenced by residential segregation. Higher levels of residential segregation play a part in the victimization–mobility relationship among Blacks in a way that is more complex than we hypothesized.

Although most crimes follow seasonal cycles, homicide is an apparent exception. The absence of homicide seasonality is surprising given that assault, a closely related offense, has an obvious annual pattern. Focusing on large U.S. cities, this article reevaluates seasonality in homicide rates using data with more extensive spatial and areal dimensions than in previous research. Panel decompositions reveal seasonal cycles in both homicide and assault rates. Seasonality stands out more clearly in assault, however, and the patterns differ somewhat in their details. The findings support the idea that assault and homicide have similar seasonal fluctuations, but they also suggest that the crimes are more distinct than criminologists often believe.

Zero-inflated Poisson (ZIP) and zero-inflated negative binomial (ZINB) models are recommended for handling excessive zeros in count data. For various reasons, researchers may not address zero inflation. This paper helps educate researchers on (1) the importance of accounting for zero inflation and (2) the consequences of misspecifying the statistical model. Using simulations, we found that when the zero inflation in the data was ignored, estimation was poor and statistically significant findings were missed. When overdispersion within the zero-inflated data was ignored, poor estimation and inflated Type I errors resulted. Recommendations on when to use the ZINB and ZIP models are provided. In an illustration using a two-step model selection procedure (likelihood ratio test and the Vuong test), the ZIP model was correctly identified only when the distributions had moderate means and sample sizes and did not correctly identify the ZINB model or the zero inflation in the ZIP and ZINB distributions.

David Greenberg has done a masterful job of considering the methodological and data requirements for a study of the New York crime drop. Anything that I could say about his presentation would amoun...

Seasonal crime patterns have been a topic of sustained criminological research for more than a century. Results in the area are often conflicting, however, and no firm consensus exists on many points. The current study uses a long time series and a large areal sample to obtain more detailed seasonality estimates than have been available in the past. The findings show that all major crime rates exhibit seasonal behavior, and that most follow similar cycles. The existence of seasonal patterns is not explainable by monthly temperature differences between areas, but seasonality and temperature variations do interact with each other. These findings imply that seasonal fluctuations have both environmental and social components, which can combine to create different patterns from one location to another.

What are "Official" RecordsThe term "official records" is broad and would include data taken from any set of administrative records in order to measure crime or features of criminal justice. In practice, it is usually taken to refer to police or court records. For present purposes, we will limit discussion to the Uniform Crime Reporting (UCR) Program. This includes the older "Summary Reporting System" with seven different data collections (Offenses Known to the Police also called Return A; Age, Sex and Race of Persons Arrested; Supplementary Homicide Report; Law Enforcement Officers Killed and Assaulted; Full-time Law Enforcement Employees; Arson Offenses Known to Law Enforcement; and Hate Crime), as well as the National Incident-Based Reporting System which is currently under development. The term "UCR" is ambiguous. When we refer to the whole organizational enterprise coordinated by the FBI, we use the term "UCR Program," otherwise we will refer to specific data collections. ... Language: en

By the time I became editor of JQC, it was already a thriving and well-regarded journal. Jamie Fox, John Laub, and Mike Maltz had done the hard work necessary to develop a niche for it within criminology, and to make its presence known to potential contributors. Due to their efforts, it had built up a solid base of readers and had earned a reputation for publishing high-quality scholarship. The editor did not have to worry about finding enough papers to fill issues, and so did not have to think much about the state of quantitative criminology in general. As with other established journals, the major editorial task was to oversee the selection of manuscripts from a pool of mostly sound and innovative submissions. These were very fortunate circumstances for me, and they reflected the significant progress that the field and the journal had made. They nevertheless provided no insights into the development of quantitative criminology, nor any illuminating stories about the role that JQC had played in it. I therefore have nothing original to say about these matters, and I will not try to disguise the fact by reflecting on them. I appreciate, however, the opportunity that the editors have given me to comment on the present state and possible future direction of quantitative criminological research. The remaining sections of this essay will first review the current situation, and will then consider what the field needs and where it may be headed.

This study examines the microlevel process of housing turnover between Blacks and Whites to assess whether crime plays an important role in the racial transition of neighborhoods. The study uses a unique, longitudinal version of the National Crime Survey in which each dwelling's close neighbors are identified. After controlling for household characteristics and the characteristics of their close neighbors, crime occurring in nearby areas is found to increase the chances of White-to-Black turnover while decreasing the chances of Black-to-White turnover. This change occurs even though the directly victimized houses do not necessarily have a probability of racial turnover different than that of other houses nearby. The findings suggest the presence of structural constraints that limit the housing opportunities for Blacks and constrain their choice of residence to comparatively unsafe neighborhoods. They also indicate that “White avoidance,” in which Whites systematically bypass high-crime neighborhoods, is important in maintaining the relationship between race and crime.

This study considers the degree to which the crime rates of US cities follow a uniform national trend. A nationwide trend has consequences for theories that explain aggregate changes in crime, but how closely subnational units hold to a common time path has received almost no research attention. Using annual panel data, the current study presents analyses that attempt to measure the correspondence between city-level and national-level crime rates. The results of each analysis are consistent with a clear single pattern that operates across the nation’s major urban areas. This supports the idea that a meaningful national trend exists, and it suggests the desirability of continuing efforts to explain it.

This article investigates the impact of criminal victimization on household residential mobility. Existing research finds that direct experiences with crime influence mobility decisions, such that persons who suffer offenses near their homes are more likely to move. The current study extends this line of inquiry to consider whether indirect victimization that involves neighbors also stimulates moving. The analysis uses the National Crime Survey to estimate multilevel models that incorporate data from individual households and their spatially proximate neighbors. The results show that the link between direct victimization and moving continues to hold after controlling for neighborhood context. Indirect property victimization also leads to moving, with effects about equal in size to those of direct victimization. In contrast, no evidence is found that violent victimization that occurs in neighboring homes influences mobility, probably because most of these events are nonstranger violence that provokes less anxiety for neighbors.

When linking records from the Supplementary Homicide Report (SHR) with information from other sources, such as data from the U.S. Census Bureau, errors are likely to occur because police service areas in some cities do not match the political definitions of the cities. Also, errors may occur because different places have the same or similar names. This article illustrates the problem with six cities (Jacksonville, Honolulu, Indianapolis, Baton Rouge, Las Vegas, and Nashville) and explains how discrepancies can be identified and corrected.

Americans move frequently, and moving alters their risks of victimization. This study uses unique longitudinal, multilevel data from the 1980–1985 National Crime Survey to examine the effects of residential turnover on household victimization. The two major findings of the study are as follows: First, housing turnover is a transition that independently increases the risk that a dwelling will experience a crime. This finding is true even controlling for persistent differences in crime vulnerability between dwellings. Second, changes in the composition and routine activities of households also alter the risks of victimization. These findings provide support for social disorganization and crime opportunity theories.

Differences between homicide estimates for large cities from 1987 through 1991 using the Supplementary Homicide Report (SHR) and the National Vital Statistics System (NVSS) are reported. Large differences exist between the NVSS and the SHR estimates expressed both as counts and as rates per 100,000 residents. Three major reasons for the differences are identified. First, many cities fail to report the SHR data in some years during the study. Second, in some places the police jurisdiction used in the SHR does not correspond to the boundaries of the city used in the NVSS. Third, the SHR associates victims with the place where the assault occurs, whereas the NVSS associates victims with the place of residence. For many purposes, the NVSS provides more accurate and appropriate measures of homicide victimization than the SHR. However, because the city-level NVSS data are requested by place of residence, they measure a different concept than the SHR, which classifies victims by place of occurrence.

Crime emerged as a topic of quantitative study in the eighteenth century and scholars today analyze it using methods drawn from all of the social sciences. Crimes are distinctive for their rarity and for their social and geographic concentration. One response to these features is to adjust for them when performing standard statistical analyses. An alternative is to take the features as essential aspects of the phenomenon under study, and give them an important role within models for longitudinal and spatial data.

This study investigated the association between victim reporting and the police response to past victimizations with data from the National Crime Victimization Survey from 1998–2000. The findings include: (1) investigatory effort by police when an individual had been victimized in the past increased the likelihood that the individual would report an ensuing victimization to the police; (2) however, this relationship only held when the victim, rather than someone else, reported the prior victimization to the police; (3) whether the police made an arrest after an individual was victimized in the past had no effect on whether the individual reported an ensuing victimization to the police; (4) the probability of victim reporting was unaffected by investigatory effort or whether an arrest was made after a prior victimization of a member of the victim’s household.