- The Ohio State University, Department of Sociology, 238 Townshend Hall, 1885
Neil Avenue Mall, Columbus, OH 43210, USA
Scott Duxbury
University of North Carolina at Chapel Hill, Sociology, Faculty Member
- Research interests: Crime, law and deviance; computational and mathematical sociology; sociology of punishment; socia... moreResearch interests: Crime, law and deviance; computational and mathematical sociology; sociology of punishment; social networks; race and ethnicity; public opinion; media and technologyedit
Mediation analysis is increasingly used in the social sciences. Extension to social network data, however, has proved difficult because statistical network models are formulated at a lower level of analysis (the dyad) than many outcomes... more
Mediation analysis is increasingly used in the social sciences. Extension to social network data, however, has proved difficult because statistical network models are formulated at a lower level of analysis (the dyad) than many outcomes of interest. This study introduces a general approach for micro-macro mediation analysis in social networks. The author defines the average mediated micro effect (AMME) as the indirect effect of a network selection process on an individual, group, or organizational outcome through its effect on an intervening network variable. The author shows that the AMME can be nonparametrically identified using a wide range of common statistical network and regression modeling strategies under the assumption of conditional independence among multiple mediators. Nonparametric and parametric algorithms are introduced to generically estimate the AMME in a multitude of research designs. The author illustrates the utility of the method with an applied example using cross-sectional National Longitudinal Study of Adolescent to Adult Health data to examine the friendship selection mechanisms that indirectly shape adolescent school performance through their effect on network structure.
Research Interests:
How do individuals' network selection decisions create unique network structures? Despite broad sociological interest in the micro-level social interactions that create macro-level network structure, few methods are available to... more
How do individuals' network selection decisions create unique network structures? Despite broad sociological interest in the micro-level social interactions that create macro-level network structure, few methods are available to statistically evaluate micro-macro relationships in social networks. This study introduces a general methodological framework for testing the effect of (micro) network selection processes, such as homophily, reciprocity, or preferential attachment, on unique (macro) network structures, such as segregation, clustering, or brokerage. The approach uses estimates from a statistical network model to decompose the contributions of each parameter to a node, subgraph, or global network statistic specified by the researcher. A flexible parametric algorithm is introduced to estimate variances, confidence intervals, and p values. Prior micro-macro network methods can be regarded as special cases of the general framework. Extensions to hypothetical network interventions, joint parameter tests, and longitudinal and multilevel network data are discussed. An example is provided analyzing the micro foundations of political segregation in a crime policy collaboration network.
Research Interests:
This centennial essay uses semi-structured topic models to trace the intellectual development of crime research in Social Forces since the 1940s.
Research Interests:
Researchers often use pooled exponential random graph models (ERGM) to analyze samples of networks. However, pooled ERGM-here, understood to include both meta-regression and combined estimation on a stacked adjacency matrix-may be biased... more
Researchers often use pooled exponential random graph models (ERGM) to analyze samples of networks. However, pooled ERGM-here, understood to include both meta-regression and combined estimation on a stacked adjacency matrix-may be biased if there is heterogeneity in the latent error variance ('scaling') of each lower-level model. This study explores the implications of scaling for pooled ERGM analysis. We illustrate that scaling can produce bias in pooled ERGM coefficients that is more severe than in single-network ERGM and we introduce two methods for reducing this bias. Simulations suggest that scaling bias can be large enough to alter conclusions about pooled ERGM coefficient size, significance, and direction, but can be substantially reduced by estimating the marginal effect within a block diagonal or random effects meta-regression framework. We illustrate each method in an empirical example using Add Health data on 15 in-school friendship networks. Results from the application illustrate that many substantive conclusions vary depending on choice of pooling method and interpretational quantity.
Research Interests:
espite a long line of scholarship on race and social control, evidence that incarceration can be connected to slavery is difficult to provide. This study evaluates whether slavery had long-term effects on growth in state incarceration... more
espite a long line of scholarship on race and social control, evidence that incarceration can be connected to slavery is difficult
to provide. This study evaluates whether slavery had long-term
effects on growth in state incarceration rates by focusing on two
key theoretical indicators: the size of the enslaved population prior
to the Civil War and demographic changes during the Great
Migration. Results reveal that states that historically had larger
enslaved populations and those that acted as destinations for
southern migrants experienced increased growth in incarceration
between 1970 and 2015. Mediation analyses show that both variables have indirect effects on change in incarceration rates through
contemporary demographic composition and public opinion. These
findings lend support to theoretical claims that link slavery to
mass incarceration and shed light on the mechanisms that enable
racial histories to map onto current criminal punishments.
to provide. This study evaluates whether slavery had long-term
effects on growth in state incarceration rates by focusing on two
key theoretical indicators: the size of the enslaved population prior
to the Civil War and demographic changes during the Great
Migration. Results reveal that states that historically had larger
enslaved populations and those that acted as destinations for
southern migrants experienced increased growth in incarceration
between 1970 and 2015. Mediation analyses show that both variables have indirect effects on change in incarceration rates through
contemporary demographic composition and public opinion. These
findings lend support to theoretical claims that link slavery to
mass incarceration and shed light on the mechanisms that enable
racial histories to map onto current criminal punishments.
Research Interests:
This study evaluates how homicide, racial threat, and media discourse interacted to shape the timing and persistence of prison growth in the United States. Drawing on Blumer's classic work, I argue that media discourse circulates threat... more
This study evaluates how homicide, racial threat, and media discourse interacted to shape the timing and persistence of prison growth in the United States. Drawing on Blumer's classic work, I argue that media discourse circulates threat narratives that portray racial minorities as either economically, politically, or criminally threatening. Criminal threat narratives increase in response to highly salient crimes, like homicide, and exert institutionally specific pressures that increase incarceration. To evaluate these claims, I use machine learning to classify 1,026,862 news articles in accordance with economic, political, and criminal threat themes in a time series analysis of the national incarceration rate between 1926 and 2016. Results reveal that the period of prison growth is characterized by an inf lux of criminal threat narratives that coincides with increases in the homicide rate. Criminal threat narratives and the homicide rate both have sizable long-term effects on the incarceration rate, whereas economic and political threat narratives have little explanatory power. Further analyses show that criminal threat narratives account for roughly half of the effect of the homicide rate on incarceration, and that the homicide rate has an indirect effect on racial disparity in prison admissions by acting through criminal threat narratives. These findings support core theoretical claims and expand our understanding of the complex interaction between racial threat and homicide in the historical rise of incarceration.
Research Interests:
Researchers often use pooled exponential random graph models (ERGM) to analyze samples of networks. However, pooled ERGM-here, understood to include both meta-regression and combined estimation on a stacked adjacency matrix-may be biased... more
Researchers often use pooled exponential random graph models (ERGM) to analyze samples of networks. However, pooled ERGM-here, understood to include both meta-regression and combined estimation on a stacked adjacency matrix-may be biased if there is heterogeneity in the latent error variance ('scaling') of each lower-level model. This study explores the implications of scaling for pooled ERGM analysis. We illustrate that scaling can produce bias in pooled ERGM coefficients that is more severe than in single-network ERGM and we introduce two methods for reducing this bias. Simulations suggest that scaling bias can be large enough to alter conclusions about pooled ERGM coefficient size, significance, and direction, but can be substantially reduced by estimating the marginal effect within a block diagonal or random effects meta-regression framework. We illustrate each method in an empirical example using Add Health data on 15 in-school friendship networks. Results from the application illustrate that many substantive conclusions vary depending on choice of pooling method and interpretational quantity.
Research Interests:
Despite decades of crime decline, police surveillance has continued to expand through a range of tactics oriented towards policing social disadvantage. Yet, despite attention to the linkages between residential inequality and policing,... more
Despite decades of crime decline, police surveillance has continued to expand through a range of tactics oriented towards policing social disadvantage. Yet, despite attention to the linkages between residential inequality and policing, few studies have accounted for two intertwined structural developments since the turn of the 21st century: (1) the shift away from spatially concentrated patterns of racial segregation within urban centers towards sprawling patterns of economic segregation and (2) the turn from reactive policing towards proactive surveillance. Using the case of big data policing, we create a new measure of big data surveillance in metropolitan areas to examine how changes in segregation have affected the expansion of proactive police surveillance. In contrast to theoretical accounts emphasizing the role of police surveillance in governing economic inequality and perpetuating racial segregation, we do not find evidence that racial segregation or income inequality incre...
Research Interests:
Although economic sociology emphasizes the role of social networks for shaping economic action, little research has examined how network governance structures affect prices in the unregulated and high-risk social context of online... more
Although economic sociology emphasizes the role of social networks for shaping economic action, little research has examined how network governance structures affect prices in the unregulated and high-risk social context of online criminal trade. We consider how overembeddedness—a state of excessive interconnectedness among market actors—arises from endogenous trade relations to shape prices in illegal online markets with aggregate consequences for short-term gross illegal revenue. Drawing on transaction-level data on 16 847 illegal drug transactions over 14 months of trade in a ‘darknet’ drug market, we assess how repeated exchanges and closure in buyer–vendor trade networks nonlinearly influence prices and short-term gross revenue from illegal drug trade. Using a series of panel models, we find that increases in closure and repeated exchange raise prices until a threshold is reached upon which prices and gross monthly revenue begin to decline as networks become overembedded. Findi...
Research Interests:
This study shows that residual variation can cause problems related to scaling in exponential random graph models (ERGM). Residual variation is likely to exist when there are unmeasured variables in a model—even those uncorrelated with... more
This study shows that residual variation can cause problems related to scaling in exponential random graph models (ERGM). Residual variation is likely to exist when there are unmeasured variables in a model—even those uncorrelated with other predictors—or when the logistic form of the model is inappropriate. As a consequence, coefficients cannot be interpreted as effect sizes or compared between models and homophily coefficients, as well as other interaction coefficients, cannot be interpreted as substantive effects in most ERGM applications. We conduct a series of simulations considering the substantive impact of these issues, revealing that realistic levels of residual variation can have large consequences for ERGM inference. A flexible methodological framework is introduced to overcome these problems. Formal tests of mediation and moderation are also proposed. These methods are applied to revisit the relationship between selective mixing and triadic closure in a large AddHealth s...
Research Interests:
Research Interests:
Research Interests:
Panel data analysis is common in the social sciences. Fixed effects models are a favorite among sociologists because they control for unobserved heterogeneity (unexplained variation) among cross-sectional units, but estimates are biased... more
Panel data analysis is common in the social sciences. Fixed effects models are a favorite among sociologists because they control for unobserved heterogeneity (unexplained variation) among cross-sectional units, but estimates are biased when there is unobserved heterogeneity in the underlying time trends. Two-way fixed effects models adjust for unobserved time heterogeneity but are inefficient, cannot include unit-invariant variables, and eliminate common trends: the portion of variance in a time-varying variable that is invariant across cross-sectional units. This article introduces a general panel model that can include unit-invariant variables, corrects for unobserved time heterogeneity, and provides the effect of common trends while also allowing for unobserved unit heterogeneity, time-varying coefficients, and time-invariant variables. One-way and two-way fixed effects models are shown to be restrictive forms of this general model. Other restrictive forms are also derived that ...
Research Interests: Sociology, Demography, Economics, Econometrics, Politics, and 11 morePanel models, Panel Data, Mass Incarceration, Econometrics Models - Panel Data, Multilevel modelling, Fixed-Effects Models, Panel Data Analysis, Mixed Models, Random-Effects Models, Sociological Methodology, and Fixed effects model
Policy responsiveness to public opinion is necessary for democratic governance, yet research on crime policy responsiveness has focused on the public at large, overlooking variation in policy support among demographic bodies and its... more
Policy responsiveness to public opinion is necessary for democratic governance, yet research on crime policy responsiveness has focused on the public at large, overlooking variation in policy support among demographic bodies and its heterogeneous effects on policy implementation. Creating new state-level measures of group opinion on punitive policy mood using 257,356 responses to 79 national surveys administered between 1970 and 2015, this study examines the effect of race, age, and gender-specific opinion on the incarceration rate. Consistent with prior studies, results reveal parallel trends in punitive policy support, with punitiveness increasing in tandem across demographic subgroups. Statistical findings from error correction models, however, reveal that not all groups had equal influence. Only the opinions of the most punitive groups—whites, men, and 30–44-year-olds—influenced the incarceration rate. Further, the largest divide in policy responsiveness to group opinion was bet...
Research Interests:
Research Interests:
Network analysis is increasingly applied throughout the social sciences. Networks have been at the core of criminological thinking since its inception. As early as Sutherland and Shaw and McKay, networks have been regarded as important... more
Network analysis is increasingly applied throughout the social sciences. Networks have been at the core of criminological thinking since its inception. As early as Sutherland and Shaw and McKay, networks have been regarded as important causes of delinquent behavior. Networks, by definition, reflect patterns of relationships between observations. Observations are typically human actors, but can represent any criminologically relevant entity, such as gangs, grocery stores, or street corners. Networks can even represent connections between actors that occupy distinct roles (e.g., connections between people and places). This flexibility in how networks can be defined and analyzed presents innumerable promising opportunities in the analysis of crime. Networks can influence criminal behavior by influencing selection into delinquent peer networks and by transmitting delinquent values and behaviors. While some of the earliest adopters of network methods turned to analyses of peer group cont...
Research Interests:
Threat theory argues that states toughen criminal laws to repress the competitive power of large minority groups. Yet, research on threat suffers from a poor understanding of why minority group size contributes to social control and a... more
Threat theory argues that states toughen criminal laws to repress the competitive power of large minority groups. Yet, research on threat suffers from a poor understanding of why minority group size contributes to social control and a lack of evidence on whether criminal law is uniquely responsive to the political interests of majority racial groups at all. By compiling a unique state-level dataset on 230 sentencing policy changes during mass incarceration and using data from 257,362 responses to 79 national surveys to construct new state-level measures of racial differences in punitive policy support, I evaluate whether criminal sentencing law is uniquely responsive to white public policy interests. Pooled event history models and mediation analyses support three primary conclusions: (1) states adopted new sentencing policies as a nonlinear response to minority group size, (2) sentencing policies were adopted in response to white public, but not black public, support for punitive c...
Research Interests:
Research Interests:
Adefining characteristic of mass incarceration is the overwhelming racial disparity in US prison populations. Yet, little research has examined the role of public attitudes in the growth of racial disparity in incarceration rates during... more
Adefining characteristic of mass incarceration is the overwhelming racial disparity in US prison populations. Yet, little research has examined the role of public attitudes in the growth of racial disparity in incarceration rates during the prison boom. This article considers the influence of explicit and modern forms of prejudice and traces a portion of the rise in racial disparity in incarceration rates to historically high levels of fear of crime. It analyzes roughly forty years of annual data on incarceration at the state level, leveraging recent methodological developments for polling data to construct longitudinal measures of state public opinion from 386,751 individual survey responses to a pooled sample of 102 national surveys. Results from panel models show that public fear of crime played a larger role in explaining rising racial disparity in incarceration rates than explicit prejudice or laissez-faire racial attitudes. Further analyses demonstrate that fear of crime media...
Research Interests:
Research Interests:
Measures of bipartite network structure have recently gained attention from network scholars. However, there is currently no measure for identifying key players in two-mode networks. This article proposes measures for identifying key... more
Measures of bipartite network structure have recently gained attention from network scholars. However, there is currently no measure for identifying key players in two-mode networks. This article proposes measures for identifying key players in bipartite networks. It focuses on two measures: fragmentation and cohesion centrality. It extends the centrality measures to bipartite networks by considering (1) cohesion and fragmentation centrality within a one-mode projection, (2) cross-modal cohesion and fragmentation centrality, where a node in one mode is influential in the one-mode projection of the other mode, and (3) cohesion and fragmentation centrality across the entire bipartite structure. Empirical examples are provided for the Southern Women’s data and on the Ndrangheta mafia data.
Research Interests:
Research Interests:
Objectives:We examine how news media portrays the causes of mass shootings for shooters of different races. Specifically, we explore whether White men are disproportionately framed as mentally ill, and what narratives media tend to invoke... more
Objectives:We examine how news media portrays the causes of mass shootings for shooters of different races. Specifically, we explore whether White men are disproportionately framed as mentally ill, and what narratives media tend to invoke when covering mass shootings through the lens of mental illness as opposed to other explanatory frames.Methods:The study examines a unique data set of 433 news documents covering 219 mass shootings between January 1, 2013, and December 31, 2015. It analyzes the data using a mixed methods approach, combining logistic regression with content analysis.Results:Quantitative findings show that Whites and Latinos are more likely to have their crime attributed to mental illness than Blacks. Qualitative findings show that rhetoric within these discussions frame White men as sympathetic characters, while Black and Latino men are treated as perpetually violent threats to the public.Conclusions:Results suggest that there is racial variability in how the media ...
Research Interests:
Exponential random graph models (ERGM) have been widely applied in the social sciences in the past ten years. However, diagnostics for ERGM have lagged behind their use. Collinearity-type problems can emerge without detection when fitting... more
Exponential random graph models (ERGM) have been widely applied in the social sciences in the past ten years. However, diagnostics for ERGM have lagged behind their use. Collinearity-type problems can emerge without detection when fitting ERGM, skewing coefficients, biasing standard errors, and yielding inconsistent model estimates. This article provides a method to detect multicollinearity in ERGM. It outlines the problem and provides a method to calculate the variance inflation factor from ERGM parameters. It then evaluates the method with a Monte Carlo simulation, fitting 216,000 ERGMs and calculating the variance inflation factors for each model. The distribution of variance inflation factors is analyzed using multilevel regression to determine what network characteristics lend themselves to collinearity-type problems. The relationship between variance inflation factors and unstable standard errors (a standard sign of collinearity) is also examined. The method is shown to effect...
Research Interests: Sociology, Computer Science, Econometrics, Statistics, Social Research Methods and Methodology, and 15 moreSocial Networks, Quantitative Methods, Social Network Analysis (SNA), Network science, Diagnostics, Exponential Random Graph, Statistical Inference on Networks, Exponential random graph models, Public health systems and services research, Sociological Methods, ERGMs, Multicollinearity, Collinearity, Monte Carlo Method, and Variance Inflation Factor
Research Interests:
Research Interests: Criminology, Computer Science, Cybercrimes, Trust, Social Network Analysis (SNA), and 14 moreCyber crime, Cybersecurity, Community Detection, Quantitative criminology, Cybercrime, Bipartite Graph, Opioid, ERGMs, Drug Markets, Online Drug Marketplaces, Drug Distribution, Co-Offending, Bipartite Networks, and Cryptomarkets
Research Interests: Business, Sociology, Criminology, Computer Science, Anthropology, and 15 moreSocial Networks, Cybercrimes, Social Media, Social Network Analysis (SNA), Criminal Networks, Social Network Analysis (Social Sciences), Illicit Drug markets, Cybercrime, Bipartite Graph, Darknet, Dark networks, ERGMs, Illegal Drug Markets, Drug Distribution, and Cryptomarkets
Potential users of grey-market substances seek out internet drug websites to gather legal high information. However, where previous researchers have investigated drug wikis as sources of drug information, few have looked into the drug... more
Potential users of grey-market substances seek out internet drug websites to gather legal high information. However, where previous researchers have investigated drug wikis as sources of drug information, few have looked into the drug forums where an abundance of legal-high information is created. Knowledge is produced on internet drug forums through social processes of drug information sharing and relating personal experiences. These knowledge production efforts are a response to internet drug forum members’ perceived need to objectively understand a drug’s behavior. This community perspective, therefore, shapes online drug forum information sharing into a marginal form of citizen science – one that does not incorporate scientist oversight or directly engage with institutional science. The article argues that drug use becomes a social ritual whereby the sharing of drug information is an ethical practice. Additionally, because first-hand experience is needed to create drug informati...
Research Interests:
How do illegal markets grow and develop? Using unique transactionlevel data on 7,205 market actors and 16,847 illegal drug exchanges on a "darknet" drug market, the authors evaluate the network processes that shape online illegal drug... more
How do illegal markets grow and develop? Using unique transactionlevel data on 7,205 market actors and 16,847 illegal drug exchanges on a "darknet" drug market, the authors evaluate the network processes that shape online illegal drug trade and promote the growth of online illegal drug markets. Contrary to past research on online markets, the authors argue that the high-risk context of illegal trade enhances market actors' reliance on social relationships that emerge endogenously from transaction networks. The findings reveal a highly structured trade network characterized by dense clustering and frequent recurrent drug exchange. Dynamic network models reveal that both embeddedness and closure in exchange structure increase the hazard rate of illegal drug trade, with effect sizes comparable to formal reputations. These effects are pronounced in the early stages of market development but wane once the market reaches maturity. These findings demonstrate the powerful, temporally contingent, influence of transaction networks on illegal trade in online markets and reveal how endogenous networks of economic relations can promote risky exchange under conditions of relative anonymity and illegality. Governments play a key role in market development. States define the type of products for sale as well as the rules governing exchange. Markets also rely on
Research Interests:
Policy responsiveness to public opinion is necessary for democratic governance, yet research on crime policy responsiveness has focused on the public at large, overlooking variation in policy support among demographic bodies and its... more
Policy responsiveness to public opinion is necessary for democratic governance, yet research on crime policy responsiveness has focused on the public at large, overlooking variation in policy support among demographic bodies and its heterogeneous effects on policy implementation. Creating new state-level measures of group opinion on punitive policy mood using 257,356 responses to 79 national surveys administered between 1970 and 2015, this study examines the effect of race, age, and gender-specific opinion on the incarceration rate. Consistent with prior studies, results reveal parallel trends in punitive policy support, with punitiveness increasing in tandem across demographic subgroups. Statistical findings from error correction models, however, reveal that not all groups had equal influence. Only the opinions of the most punitive groups—whites, men, and 30 – 44 year olds—influenced the incarceration rate. Further, the largest divide in policy responsiveness to group opinion was between blacks and whites. In fact, estimates suggest that if whites’ punitive policy support had not risen, the current incarceration rate would be roughly the same as if violent crime had not risen during the 45 years of mass incarceration. The implications of these results are discussed for research on policy responsiveness, partisan politics, and for the “democracy at work” hypothesis of mass incarceration.
Research Interests:
Although economic sociology emphasizes the role of social networks for shaping economic action, little research has examined how network governance structures affect prices in the unregulated and high-risk social context of online... more
Although economic sociology emphasizes the role of social networks for shaping economic action, little research has examined how network governance structures affect prices in the unregulated and high-risk social context of online criminal trade. We consider how overembeddedness-a state of excessive interconnectedness among market actors-arises from endogenous trade relations to shape prices in illegal online markets with aggregate consequences for short-term gross illegal revenue. Drawing on transaction-level data on 16 847 illegal drug transactions over 14 months of trade in a 'darknet' drug market, we assess how repeated exchanges and closure in buyervendor trade networks nonlinearly influence prices and short-term gross revenue from illegal drug trade. Using a series of panel models, we find that increases in closure and repeated exchange raise prices until a threshold is reached upon which prices and gross monthly revenue begin to decline as networks become overembedded. Findings provide insight into the network determinants of prices and gross monthly revenue in illegal online drug trade and illustrate how network structure shapes prices in criminal markets, even in anonymous trade environments.
Research Interests:
Panel data analysis is common in the social sciences. Fixed effects models are a favorite among sociologists because they control for unobserved heterogeneity (unexplained variation) among cross-sectional units, but estimates are biased... more
Panel data analysis is common in the social sciences. Fixed effects models are a favorite among sociologists because they control for unobserved heterogeneity (unexplained variation) among cross-sectional units, but estimates are biased when there is unobserved heterogeneity in the underlying time trends. Twoway fixed effects models adjust for unobserved time heterogeneity but are inefficient, cannot include unitinvariant variables, and eliminate common trends: the portion of variance in a time-varying variable that is invariant across cross-sectional units. This article introduces a general panel model that can include unitinvariant variables, corrects for unobserved time heterogeneity, and provides the effect of common trends while also allowing for unobserved unit heterogeneity, time-varying coefficients, and time-invariant variables. One-way and two-way fixed effects models are shown to be restrictive forms of this general model. Other restrictive forms are also derived that offer all the usual advantages of one-way and two-way fixed effects models but account for unobserved time heterogeneity. The author uses the models to examine the increase in state incarceration rates between 1970 and 2015.
Research Interests:
This study shows that residual variation can cause problems related to scaling in exponential random graph models (ERGM). Residual variation is likely to exist when there are unmeasured variables in a model-even those uncor-related with... more
This study shows that residual variation can cause problems related to scaling in exponential random graph models (ERGM). Residual variation is likely to exist when there are unmeasured variables in a model-even those uncor-related with other predictors-or when the logistic form of the model is inappropriate. As a consequence, coefficients cannot be interpreted as effect sizes or compared between models and homophily coefficients, as well as other interaction coefficients, cannot be interpreted as substantive effects in most ERGM applications. We conduct a series of simulations considering the substantive impact of these issues, revealing that realistic levels of residual variation can have large consequences for ERGM inference. A flexible methodological framework is introduced to overcome these problems. Formal tests of mediation and moderation are also proposed. These methods are applied to revisit the relationship between selective mixing and triadic closure in a large AddHealth school friendship network. Extensions to other classes of statistical work models are discussed.
Research Interests:
A defining characteristic of mass incarceration is the overwhelming racial disparity in US prison populations. Yet, little research has examined the role of public attitudes in the growth of racial disparity in incarceration rates during... more
A defining characteristic of mass incarceration is the overwhelming racial disparity in US prison populations. Yet, little research has examined the role of public attitudes in the growth of racial disparity in incarceration rates during the prison boom. This article considers the influence of explicit and modern forms of prejudice and traces a portion of the rise in racial disparity in incarceration rates to historically high levels of fear of crime. It analyzes roughly forty years of annual data on incarceration at the state level, leveraging recent methodological developments for polling data to construct longitudinal measures of state public opinion from 386,751 individual survey responses to a pooled sample of 102 national surveys. Results from panel models show that public fear of crime played a larger role in explaining rising racial disparity in incarceration rates than explicit prejudice or laissez-faire racial attitudes. Further analyses demonstrate that fear of crime mediates a sizable portion of the effect of race-specific offending. These findings reveal that it is fear of crime and its radicalized overtones, rather than explicit or laissez-faire racial attitudes, that contributes to much of the observed racial inequality in the criminal justice system. The racial disparity in incarceration rates has proven to be "one of the most distressing and troublesome aspects of the criminal justice system" (Blumstein 1982, 1,259). With a black incarceration rate six times larger than the white incarceration rate (Beck and Blumstein 2018), the demographic composition of US prisons boasts a level of racial disparity greater than comparable disparities in wealth, employment, or education (Western 2007). Criminal records follow prisoners across the life-course and bar them from opportunities for upward
Research Interests:
Threat theory argues that states toughen criminal laws to repress the competitive power of large minority groups. Yet, research on threat suffers from a poor understanding of why minority group size contributes to social control and a... more
Threat theory argues that states toughen criminal laws to repress the competitive power of large minority groups. Yet, research on threat suffers from a poor understanding of why minority group size contributes to social control and a lack of evidence on whether criminal law is uniquely responsive to the political interests of majority racial groups at all. By compiling a unique state-level dataset on 230 sentencing policy changes during mass incarceration and using data from 257,362 responses to 79 national surveys to construct new state-level measures of racial differences in punitive policy support, I evaluate whether criminal sentencing law is uniquely responsive to white public policy interests. Pooled event history models and mediation analyses support three primary conclusions: (1) states adopted new
sentencing policies as a nonlinear response to minority group size, (2) sentencing policies were adopted in response to white public, but not black public, support for punitive crime policy, and (3) minority group size and race-specific homicide victimization both indirectly affect sentencing policy by increasing white public punitive policy support. These findings support key theoretical propositions for the threat explanation of legal change and identify white public policy opinion as a mechanism linking minority group size to variation in criminal law.
sentencing policies as a nonlinear response to minority group size, (2) sentencing policies were adopted in response to white public, but not black public, support for punitive crime policy, and (3) minority group size and race-specific homicide victimization both indirectly affect sentencing policy by increasing white public punitive policy support. These findings support key theoretical propositions for the threat explanation of legal change and identify white public policy opinion as a mechanism linking minority group size to variation in criminal law.
Research Interests:
Physical, technological, and social networks are often at risk of intentional attack. Despite the wide-spanning importance of network vulnerability, very little is known about how criminal networks respond to attacks or whether... more
Physical, technological, and social networks are often at risk of intentional attack. Despite the wide-spanning importance of network vulnerability, very little is known about how criminal networks respond to attacks or whether intentional attacks affect criminal activity in the long-run. To assess criminal network responsiveness, we designed an empirically-grounded agent-based simulation using population-level network data on 16,847 illicit drug exchanges between 7,295 users of an active darknet drug market and statistical methods for simulation analysis. We consider three attack strategies: targeted attacks that delete structurally integral vertices, weak link attacks that delete large numbers of weakly connected vertices, and signal attacks that saturate the network with noisy signals. Results reveal that, while targeted attacks are effective when conducted at a large-scale, weak link and signal attacks deter more potential drug transactions and buyers when only a small portion of the network is attacked. We also find that intentional attacks affect network behavior. When networks are attacked, actors grow more cautious about forging ties, connecting less frequently and only to trustworthy alters. Operating in tandem, these two processes undermine long-term network robustness and increase network vulnerability to future attacks.
Research Interests:
Legitimacy is widely invoked as a master frame in international political discourse. During episodes of contention, this frame is used by opposing sides to advance competing interpretations of the same social problems. Through an analysis... more
Legitimacy is widely invoked as a master frame in international political discourse. During episodes of contention, this frame is used by opposing sides to advance competing interpretations of the same social problems. Through an analysis of elite political discourses surrounding international intervention in the Syrian Civil War, we examine what distinguishes the effectiveness of actors' framing efforts when they use a shared frame to advance conflicting agendas. We show how features of the objects (i.e., what or who) being framed shape the resonance and stability of the framing. Moreover, we show how framing objects that can be coherently interpreted in multiple ways facilitate the cultivation of discourses that are consistent despite changing social conditions and the evolution of framers' goals. We refer to this as robust discourse and elaborate on the implications of this concept.
Measures of bipartite network structure have recently gained attention from network scholars. However, there is currently no measure for identifying key players in two-mode networks. This article proposes measures for identifying key... more
Measures of bipartite network structure have recently gained attention from network scholars. However, there is currently no measure for identifying key players in two-mode networks. This article proposes measures for identifying key players in bipartite networks. It focuses on two measures: fragmentation and cohesion centrality. It extends the centrality measures to bipartite networks by considering (1) cohesion and fragmentation centrality within a one-mode projection, (2) cross-modal cohesion and fragmentation centrality, where a node in one mode is influential in the one-mode projection of the other mode, and (3) cohesion and fragmentation centrality across the entire bipartite structure. Empirical examples are provided for the Southern Women's data and on the Ndrangheta mafia data.
Objectives: We examine how news media portrays the causes of mass shootings for shooters of different races. Specifically, we explore whether White men are disproportionately framed as mentally ill, and what narratives media tend to... more
Objectives: We examine how news media portrays the causes of mass shootings for shooters of different races. Specifically, we explore whether White men are disproportionately framed as mentally ill, and what narratives media tend to invoke when covering mass shootings through the lens of mental illness as opposed to other explanatory frames. Methods: The study examines a unique data set of 433 news documents covering 219 mass shootings between January 1, 2013, and December 31, 2015. It analyzes the data using a mixed methods approach, combining logistic regression with content analysis. Results: Quantitative findings show that Whites and Latinos are more likely to have their crime attributed to mental illness than Blacks. Qualitative findings show that rhetoric within these discussions frame White men as sympathetic characters, while Black
Research Interests:
Exponential random graph models (ERGM) have been widely applied in the social sciences in the past ten years. However, diagnostics for ERGM have lagged behind their use. Collinearity-type problems can emerge without detection when fitting... more
Exponential random graph models (ERGM) have been widely applied in the social sciences in the past ten years. However, diagnostics for ERGM have lagged behind their use. Collinearity-type problems can emerge without detection when fitting ERGM, yielding inconsistent model estimates and problematizing inference from parameters. This article provides a method to detect multicollinearity in ERGM. It outlines the problem and provides a method to calculate the variance inflation factor from ERGM parameters. It then evaluates the method with a Monte Carlo simulation, fitting 216,000 ERGMs and calculating the variance inflation factors for each model. The distribution of variance inflation factors is analyzed using multilevel regression to determine what network characteristics lend themselves to collinearity-type problems. The relationship between variance inflation factors and unstable standard errors (a standard sign of collinearity) is also examined. The method is shown to effectively detect multicollinearity and guidelines for interpretation are discussed.
Research Interests:
Drug distributors are increasingly turning to online markets to deliver and procure illegal drugs. Online venues allow drug vendors to span broad audiences, reshape organizational structure, and remain relatively anonymous. Such trends... more
Drug distributors are increasingly turning to online markets to deliver and procure illegal drugs. Online venues allow drug vendors to span broad audiences, reshape organizational structure, and remain relatively anonymous. Such trends raise fundamental questions regarding the structural robustness, topological characteristics, and tie formation patterns in online drug distribution networks. We examine one online illegal opioid transaction network. We characterize the network’s topology, evaluate selection dynamics that sustain and facilitate the growth of the drug market, and investigate network vulnerability. Results support the existence of trust-based preferential attachment and give insight to how the network reacts to disruption.
Research Interests: Sociology, Criminology, Social Networks, Cybercrimes, Social Media, and 12 moreSocial Network Analysis (SNA), Criminal Networks, Social Network Analysis (Social Sciences), Illicit Drug markets, Cybercrime, Bipartite Graph, Darknet, Dark networks, ERGMs, Illegal Drug Markets, Drug Distribution, and Cryptomarkets
Objectives: The current study is the first to examine the network structure of an encrypted online drug distribution network. It examines 1) the global network structure, 2) the local network structure, and 3) identifies those vendor... more
Objectives: The current study is the first to examine the network structure of an encrypted online drug distribution network. It examines 1) the global network structure, 2) the local network structure, and 3) identifies those vendor characteristics that best explain variation in the network structure. In doing so, it evaluates the role of trust in online drug markets.
Methods: The study draws on a unique dataset of transaction level data from an encrypted online drug market. Structural measures and community detection analysis are used to characterize and investigate the network structure. Exponential random graph modeling is used to evaluate which vendor characteristics explain variation in purchasing patterns.
Results: Vendors’ trustworthiness explains more variation in the overall network structure than the affordability of vendor products or the diversity of vendor product listings. This results in a highly localized network structure with a few key vendors accounting for most transactions.
Conclusions: The results indicate that vendors’ trustworthiness is a better predictor of vendor selection than product diversity or affordability. These results illuminate the internal market dynamics that sustain digital drug markets and highlight the importance of examining how new anonymizing technologies shape global drug distribution networks.
Methods: The study draws on a unique dataset of transaction level data from an encrypted online drug market. Structural measures and community detection analysis are used to characterize and investigate the network structure. Exponential random graph modeling is used to evaluate which vendor characteristics explain variation in purchasing patterns.
Results: Vendors’ trustworthiness explains more variation in the overall network structure than the affordability of vendor products or the diversity of vendor product listings. This results in a highly localized network structure with a few key vendors accounting for most transactions.
Conclusions: The results indicate that vendors’ trustworthiness is a better predictor of vendor selection than product diversity or affordability. These results illuminate the internal market dynamics that sustain digital drug markets and highlight the importance of examining how new anonymizing technologies shape global drug distribution networks.
Research Interests: Cybercrimes, Trust, Social Network Analysis (SNA), Cyber crime, Cybersecurity, and 11 moreOrganized Crime, Drugs and Crime, Cybercrime, Community Detection, Cybercrime, Bipartite Graph, ERGMs, Drug Markets, Online Drug Marketplaces, Drug Distribution, Co-Offending, Bipartite Networks, and Cryptomarkets
Although longitudinal social network data are increasingly collected, there are few guides on how to navigate the range of available tools for longitudinal network analysis. The applied social scientist is left to wonder: Which model is... more
Although longitudinal social network data are increasingly collected, there are few guides on how to navigate the range of available tools for longitudinal network analysis. The applied social scientist is left to wonder: Which model is most appropriate for my data? How should I get started with this modeling strategy? And how do I know if my model is any good? This book answers these questions. Author Scott Duxbury assumes that the reader is familiar with network measurement, description, and notation, and is versed in regression analysis, but is likely unfamiliar with statistical network methods. The goal of the book is to guide readers towards choosing, applying, assessing, and interpreting a longitudinal network model, and each chapter is organized with a specific data structure or research question in mind. A companion website includes data and R