Hiding the Hate—Contextual Effects on Hate Crime Reports

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is an interesting paper with the potential to make contributions to the field of research on hate crimes and their underpinnings. The reasons for rating the paper “Average” are as follows.

1. The promise as expressed under sections 1 and 2 are not realized in the presentation of the results. Instead, section 6 is primarily focused on showing statistical techniques rather than showing the importance of the findings for the research questions.

2. Section 7 the discussion section functions as a summary/recap of the paper and fails to engage with the either the theoretical or empirical position taken at the beginning of the paper.

3. Importantly, the paper lacks readability. Specifically, section 6 and 7 require more attention. A focus on displaying statistical piousness, while admirable, means that the papers contributions may not meet the broader audience it deserves.

My suggestions:

The statistics are necessary. However, I suggest that the Author(s) rethink which are necessary and which are peripheral – can the same results be achieved with fewer statistics. This would leave space for presenting results using more words. Finding the right balance between statistical presentations and lay explanations of what they mean for the research questions is key.

Secondly, the Author(s) need to link the results more clearly to the presentations under sections 1 and 2. For example, what new knowledge has this study contributed? What theoretical position does it question, which debates does it participate in?

Thirdly, I suggest that the Author(s) consider their presumptive audience – who is the target reader? By rethinking this question, the Author(s) can revise with focus on improving readability.

Author Response

This is an interesting paper with the potential to make contributions to the field of research on hate crimes and their underpinnings. The reasons for rating the paper “Average” are as follows.

1. The promise as expressed under sections 1 and 2 are not realized in the presentation of the results. Instead, section 6 is primarily focused on showing statistical techniques rather than showing the importance of the findings for the research questions.
2. Section 7 the discussion section functions as a summary/recap of the paper and fails to engage with the either the theoretical or empirical position taken at the beginning of the paper.
3. Importantly, the paper lacks readability. Specifically, section 6 and 7 require more attention. A focus on displaying statistical piousness, while admirable, means that the papers contributions may not meet the broader audience it deserves.

My suggestions:

The statistics are necessary. However, I suggest that the Author(s) rethink which are necessary and which are peripheral – can the same results be achieved with fewer statistics. This would leave space for presenting results using more words. Finding the right balance between statistical presentations and lay explanations of what they mean for the research questions is key.

Reply: Thank you for this valuable feedback. I appreciate your suggestion to reconsider the balance between statistical presentations and textual explanations.

After careful consideration, I have decided to retain the current statistical analyses. These statistics are an essential part of the empirical underpinning of the various assumptions discussed in the text. They provide crucial evidence of the robustness and validity of the results. The decision to retain these statistics is also based on positive feedback from other reviewers, including reviewer 2, whose feedback led to additional statistical analysis. I have taken this feedback into account by only briefly mentioning the additional statistics in the main text and including the detailed results in the appendix.

To address the concern about the balance between statistical detail and interpretative discussion, I have made efforts to enhance the explanatory sections. This should help in making the findings more accessible while preserving the necessary statistical rigor.

I hope this approach aligns with the reviewer's suggestions and improves the overall readability and impact of the paper.

 

Secondly, the Author(s) need to link the results more clearly to the presentations under sections 1 and 2. For example, what new knowledge has this study contributed? What theoretical position does it question, which debates does it participate in?

Reply: Reply: Thank you for the comment, I have now more explicitly linked the results with the theoretical preliminary considerations.

Thirdly, I suggest that the Author(s) consider their presumptive audience – who is the target reader? By rethinking this question, the Author(s) can revise with focus on improving readability.

Reply: Thank you for your comment. I agree that the readability of a text depends very much on the target audience. The presumed target audience for this manuscript is interested academics who have already studied or have a strong interest in the topic of hate crime and hate crime reporting, and thus have some prior knowledge in this area. Also with reference to the otherwise very positive feedback from the other three reviewers, I have deliberately kept the text at a level that assumes this level of knowledge. Nevertheless, I have tried to take the feedback into account and see if there are ways to further improve the readability without reducing the depth and complexity of the subject matter.

Reviewer 2 Report

Comments and Suggestions for Authors

In an otherwise ambitious analytical work, I do not find much if any consideration of the possibility that hate crime reporting by local police might be affected by the fact that local police indeed reflect the local political culture in which they are operating. Thus, conservative, white privileged communities are likely to be policed by conservative, white privileged police departments.  Those police enforce or not, as the case might be, what the community wants enforced or not!  This might mean that certain, what in other communities might be hate crimes, are more acceptable in such communities.

 

Do white supremacist views and norms in a conservative society mean there is less concern in that society about basic civil rights – particularly the rights of minority non-whites?  And, is this lesser concern reflected in less attention to violations of those civil rights (including in the form of hate crimes) by local police as shown by fewer police reports of such violations?   Addressing these questions requires multiple, interconnected operational measures, and the study’s conclusions are only as good as the measures and their analyses used to reach them.  The author also has to make a number of assumptions, which then have to be seen as being reasonable in order for one to trust the results. 

 

 

In general, the author here has done a good job of presenting the research questions and hypotheses, and of explaining and justifying the various measures that are used.  There are, however, several issues to which I recommend the author give some further attention.  First, it should be more strongly acknowledged that by choosing white supremacy and white privilege as the societal context for hate crimes, one is sidestepping or downplaying the point that there are, as the author recognizes, other forms of hate crimes not based upon race, but rather on gender, sexual orientation, religion, physical disability, etc.  Persons in those categories also have civil rights that are violated by hate groups, not necessarily motivated by racial hatred and/or white supremacy.  Second, I understand the rationale for choosing hate crime reports as opposed to actual hate crimes as a dependent variable, but it should be recognized that the limitations mentioned about the actual crime data should cut across all the counties studied. Thus, these data would be comparable.  Thirdly, I recommend the author give further thought to the idea that the police – all police – largely represent the power structures of the communities they are policing.  They enforce against those behaviors that the influential members of the community want enforced and vice versa.  Thus, if in a particular community the dominant political culture is one of white privilege and supremacy, it would not be surprising that the police enforce the law accordingly.  Irrespective of the actual number of racially motivated hate crimes, might one not expect to see fewer police reports? 

 

Author Response

In an otherwise ambitious analytical work, I do not find much if any consideration of the possibility that hate crime reporting by local police might be affected by the fact that local police indeed reflect the local political culture in which they are operating. Thus, conservative, white privileged communities are likely to be policed by conservative, white privileged police departments.  Those police enforce or not, as the case might be, what the community wants enforced or not!  This might mean that certain, what in other communities might be hate crimes, are more acceptable in such communities.

Do white supremacist views and norms in a conservative society mean there is less concern in that society about basic civil rights – particularly the rights of minority non-whites?  And, is this lesser concern reflected in less attention to violations of those civil rights (including in the form of hate crimes) by local police as shown by fewer police reports of such violations?   Addressing these questions requires multiple, interconnected operational measures, and the study’s conclusions are only as good as the measures and their analyses used to reach them.  The author also has to make a number of assumptions, which then have to be seen as being reasonable in order for one to trust the results.   

In general, the author here has done a good job of presenting the research questions and hypotheses, and of explaining and justifying the various measures that are used.  There are, however, several issues to which I recommend the author give some further attention.  First, it should be more strongly acknowledged that by choosing white supremacy and white privilege as the societal context for hate crimes, one is sidestepping or downplaying the point that there are, as the author recognizes, other forms of hate crimes not based upon race, but rather on gender, sexual orientation, religion, physical disability, etc.  Persons in those categories also have civil rights that are violated by hate groups, not necessarily motivated by racial hatred and/or white supremacy. 

Reply: Thank you for this insightful comment. I fully agree with the reviewer that not all hate crimes are necessarily motivated by white supremacy or white privilege, and that hate crimes based on gender, sexual orientation, religion, physical disability, etc., should also be recognized as serious violations of civil rights.

To address this, I propose a broader interpretation of white supremacy and white privilege. While these concepts primarily refer to a racist societal perspective, they also correlate strongly with the devaluation of other social groups, such as those with non-traditional sexual orientations or non-conforming gender identities. In other words, individuals who exhibit preferences for white supremacy or white privilege often hold prejudiced views against various marginalized groups, not just along racial lines.

Empirically, I approached this issue by including only incidents where the perpetrator was either white or of unknown race. I chose to include unknown perpetrators because the literature suggests that hate crime offenders are often male and white. However, this does not exclude the possibility that some perpetrators in the final sample may not be white.

As an additional robustness check, I conducted separate analyses for incidents involving only white perpetrators, only unknown perpetrators, and only non-white perpetrators. These analyses controlled for the interaction between hate groups and the current Republican vote share, alongside other control variables used in the main text. The general dynamics appear comparable when looking at the significant effects across these models. Notably, the interaction between hate groups and Republican vote share shows a significant effect similar to the main findings, but this effect is absent in the case of non-white perpetrators. This suggests that different social mechanisms may be at play, and that the mechanisms involving white and unknown perpetrators are comparable.

I have added these models and a brief explanation in the appendix, with a corresponding reference in the main text.

Second, I understand the rationale for choosing hate crime reports as opposed to actual hate crimes as a dependent variable, but it should be recognized that the limitations mentioned about the actual crime data should cut across all the counties studied. Thus, these data would be comparable. 

Reply: Thank you for your comment, I appreciate the point about the comparability of actual hate crime data across countries. However, I would like to clarify and reinforce my reasons for focusing on hate crime reports:

1. Differences in reporting and underreporting: While it may be true that the limitations of actual hate crime data may be consistent across counties, it is equally plausible that the problem of underreporting varies significantly depending on factors such as community trust in law enforcement, local law enforcement practices and social stigma. These variations can lead to inconsistent and potentially biased data when using actual hate crime as a measure. By focusing on hate crime reporting, I aim to capture not only the incidents themselves, but also the social dynamics that influence the reporting of these crimes.
2. Reflecting societal responses: Hate crime reports provide an insight into societal responses to hate crime, which is an important aspect of understanding the wider context of hate-related incidents. The decision to report a hate crime is influenced by a number of factors, including the perceived seriousness of the incident, community support and the effectiveness of local law enforcement. This makes hate crime reporting a more nuanced indicator that is more in line with the aims of the study.
3. Consistency with research objectives: The research aims to explore not only the occurrence of hate crimes, but also the mechanisms that influence their reporting. By using hate crime reports as the dependent variable, the study can better account for and analyse the social, cultural and institutional factors that contribute to how hate crimes are recognised and addressed within communities.

For these reasons, I believe that hate crime reports are a more appropriate dependent variable for this study, as they provide a richer and more contextual understanding of the phenomena under investigation. I hope this explanation clarifies my approach and look forward to your further thoughts.

 Thirdly, I recommend the author give further thought to the idea that the police – all police – largely represent the power structures of the communities they are policing.  They enforce against those behaviors that the influential members of the community want enforced and vice versa.  Thus, if in a particular community the dominant political culture is one of white privilege and supremacy, it would not be surprising that the police enforce the law accordingly.  Irrespective of the actual number of racially motivated hate crimes, might one not expect to see fewer police reports? 

Reply: Thanks for the advice! I have tried to address this issue in the discussion. Unfortunately, it is not possible to draw direct conclusions about these social dynamics from the available data, but I find the reviewer's argument very plausible. Especially given the direct election of county sheriffs by the local population and their sometimes extremely conservative (by US standards) political orientation, it seems likely that this is reflected in their daily policing practices.

Reviewer 3 Report

Comments and Suggestions for Authors

This is a paper that requires very little in advice to the author. The paper describes an excellent research project using best practices in crtitical analysis and strong statistically analysed data statistics  to address in a nuanced manner  key hypotheses relating the intensity of hate crime reporting in the US with multiple contextual factors.

The findings are carefully deduced and explained without attributing blame or malfeasance of any sort by any of the actors.

After setting the scene of the research in  terms of the existing literature, the author critically points to the need for research in this area to refocus on the *rate of recording* of hate crime incidents rather than than the rate of hate crime incidents. This brings the focus onto how sundry factors influence whether and when hate crimes are recorded.

The author identifies and addresses the limitations of the data, the analyses and the findings.

In short this is an excellent paper that could be published as is.

The only changes I would recommend are the change to 'recording and enforcement' in line 401 and that references are missing from line 465. There are also a small number of other points in the discussion that would benefit from pointers to appropriate references but these latter are not essential.

 

Author Response

Review 3

This is a paper that requires very little in advice to the author. The paper describes an excellent research project using best practices in crtitical analysis and strong statistically analysed data statistics  to address in a nuanced manner  key hypotheses relating the intensity of hate crime reporting in the US with multiple contextual factors.

The findings are carefully deduced and explained without attributing blame or malfeasance of any sort by any of the actors.

After setting the scene of the research in  terms of the existing literature, the author critically points to the need for research in this area to refocus on the *rate of recording* of hate crime incidents rather than than the rate of hate crime incidents. This brings the focus onto how sundry factors influence whether and when hate crimes are recorded.

The author identifies and addresses the limitations of the data, the analyses and the findings.

In short this is an excellent paper that could be published as is.

The only changes I would recommend are the change to 'recording and enforcement' in line 401 and that references are missing from line 465. There are also a small number of other points in the discussion that would benefit from pointers to appropriate references but these latter are not essential.

Reply: Thank you very much for this overall very positive feedback! I have adopted the suggestions in the manuscript and changed the aspects you requested.

Reviewer 4 Report

Comments and Suggestions for Authors

The manuscript titled "Hiding the Hate – Contextual Effects on Hate Crime Reports" delves into the intricate dynamics of hate crime reporting within the United States, particularly emphasizing the impact of local norm shifts influenced by white supremacist ideologies. The study utilizes a comprehensive dataset spanning from 2010 to 2020, incorporating data from the FBI UCR, SPLC, MIT Election Lab, and the U.S. Census, to explore how the activity of hate groups and political affiliations at the county level affect the reporting of hate crimes.

The abstract succinctly summarizes the findings, highlighting a threefold increase in expected hate crime reports in counties with heightened hate group activity. However, this effect is moderated by Republican vote share, which appears to reduce reported hate crimes by 23% in counties with strong Republican support and significant right-wing hate group presence. Additionally, adjacent Republican counties also exhibit a spillover effect, with a 13% reduction in expected hate crime reports. Interestingly, the regional impact is also considered, noting that Northeastern counties with higher right-wing hate group presence show a 23% lower incidence rate of reported hate crimes.

The introduction sets the stage by contextualizing hate crimes as message crimes that reinforce power structures within communities. It traces the historical evolution of hate crime legislation and reporting mechanisms, emphasizing the discrepancies in hate crime data between the FBI's UCR and the NCVS, which suggest significant underreporting by law enforcement agencies. This discrepancy forms the basis of the study’s hypothesis that hate crime reports reflect the degree to which local law enforcement enforces civil rights rather than the actual occurrence of bias-motivated behavior.

The theoretical framework and empirical background sections provide a detailed review of existing literature on the influence of local social, cultural, and political contexts on hate crime reporting. The study posits that white supremacist norms at the county level detrimentally affect the enforcement of hate crime reporting. The discussion extends to right-wing hate groups, used as proxies for white supremacist attitudes, and their spatial and regional distribution, with a focus on their influence on local social norms and political processes.

The data section outlines the various sources and methods used to compile the dataset, addressing potential biases and ensuring robustness through multiple regression models. The dependent variable is defined as officially reported hate crimes, with a careful exclusion of anti-white and anti-male hate crimes to maintain focus on groups structurally targeted by white supremacist hostility. Independent variables include the presence of active hate groups and political orientation measured by Republican vote share, among other socioeconomic and demographic factors.

The results section presents the findings through incidence rate ratios (IRR) from negative binomial regressions. The analysis reveals no general effect of locally active white supremacist hate groups on hate crime reports but identifies significant spillover effects on neighboring counties. The interaction effects between hate groups and Republican vote share suggest a nuanced relationship, where conservative counties show lower expected hate crime reports. Additionally, regional differences highlight varying impacts, with Northeastern counties exhibiting lower hate crime reports in the presence of active hate groups.

In the discussion, the study emphasizes the complexity of the relationship between local contexts, hate groups, and law enforcement reporting behavior. It suggests that reduced enforcement of hate crime reporting in certain areas may be an unintended consequence of local norm shifts influenced by white supremacist ideologies. The paper calls for a more nuanced understanding of broader social mechanisms affecting hate crime reporting and the need for further research to address the underlying social dynamics.

The manuscript concludes with a critical reflection on the implications of the findings, acknowledging limitations and proposing directions for future research. It underscores the importance of promoting the enforcement of civil rights and deconstructing white supremacist and white privilege norms, especially as the U.S. becomes more racially diverse.

Overall, the study offers a comprehensive and insightful analysis of the contextual effects on hate crime reporting, contributing significantly to the understanding of how local social and political factors influence law enforcement behavior and the broader enforcement of civil rights.

Author Response

Review 4

The manuscript titled "Hiding the Hate – Contextual Effects on Hate Crime Reports" delves into the intricate dynamics of hate crime reporting within the United States, particularly emphasizing the impact of local norm shifts influenced by white supremacist ideologies. The study utilizes a comprehensive dataset spanning from 2010 to 2020, incorporating data from the FBI UCR, SPLC, MIT Election Lab, and the U.S. Census, to explore how the activity of hate groups and political affiliations at the county level affect the reporting of hate crimes.

The abstract succinctly summarizes the findings, highlighting a threefold increase in expected hate crime reports in counties with heightened hate group activity. However, this effect is moderated by Republican vote share, which appears to reduce reported hate crimes by 23% in counties with strong Republican support and significant right-wing hate group presence. Additionally, adjacent Republican counties also exhibit a spillover effect, with a 13% reduction in expected hate crime reports. Interestingly, the regional impact is also considered, noting that Northeastern counties with higher right-wing hate group presence show a 23% lower incidence rate of reported hate crimes.

The introduction sets the stage by contextualizing hate crimes as message crimes that reinforce power structures within communities. It traces the historical evolution of hate crime legislation and reporting mechanisms, emphasizing the discrepancies in hate crime data between the FBI's UCR and the NCVS, which suggest significant underreporting by law enforcement agencies. This discrepancy forms the basis of the study’s hypothesis that hate crime reports reflect the degree to which local law enforcement enforces civil rights rather than the actual occurrence of bias-motivated behavior.

The theoretical framework and empirical background sections provide a detailed review of existing literature on the influence of local social, cultural, and political contexts on hate crime reporting. The study posits that white supremacist norms at the county level detrimentally affect the enforcement of hate crime reporting. The discussion extends to right-wing hate groups, used as proxies for white supremacist attitudes, and their spatial and regional distribution, with a focus on their influence on local social norms and political processes.

The data section outlines the various sources and methods used to compile the dataset, addressing potential biases and ensuring robustness through multiple regression models. The dependent variable is defined as officially reported hate crimes, with a careful exclusion of anti-white and anti-male hate crimes to maintain focus on groups structurally targeted by white supremacist hostility. Independent variables include the presence of active hate groups and political orientation measured by Republican vote share, among other socioeconomic and demographic factors.

The results section presents the findings through incidence rate ratios (IRR) from negative binomial regressions. The analysis reveals no general effect of locally active white supremacist hate groups on hate crime reports but identifies significant spillover effects on neighboring counties. The interaction effects between hate groups and Republican vote share suggest a nuanced relationship, where conservative counties show lower expected hate crime reports. Additionally, regional differences highlight varying impacts, with Northeastern counties exhibiting lower hate crime reports in the presence of active hate groups.

In the discussion, the study emphasizes the complexity of the relationship between local contexts, hate groups, and law enforcement reporting behavior. It suggests that reduced enforcement of hate crime reporting in certain areas may be an unintended consequence of local norm shifts influenced by white supremacist ideologies. The paper calls for a more nuanced understanding of broader social mechanisms affecting hate crime reporting and the need for further research to address the underlying social dynamics.

The manuscript concludes with a critical reflection on the implications of the findings, acknowledging limitations and proposing directions for future research. It underscores the importance of promoting the enforcement of civil rights and deconstructing white supremacist and white privilege norms, especially as the U.S. becomes more racially diverse.

Overall, the study offers a comprehensive and insightful analysis of the contextual effects on hate crime reporting, contributing significantly to the understanding of how local social and political factors influence law enforcement behavior and the broader enforcement of civil rights.

Reply: Thank you very much for the positive feedback! I could not find any suggestions for changes and have therefore not made any further changes to the manuscript.