Decoding the U.S. justice system’s response to typologies of domestic terrorism

This study adjudicates disparities in the United States government’s response to different terrorism typologies under a domestic legal framework. Far Right and Islamic Extremist typologies were compared and informed by Structural-Contextual theory. The study leveraged data from the American Terrorism Study to evaluate empirically prosecutorial approaches, plea and trial conviction rates, the magnitude of investigatory resources applied, and levels of explained variance between groups. Total conviction rates were largely different among groups, with significantly harsher outcomes for Islamic Extremists. The author proposes a variation of Structural-Contextual theory, where a proactive political environment is predictive of more severe outcomes.


Introduction
Americans regard the pursuit of justice as a fundamental government process, intended both to protect the populace and punish offenders. They expect it to be approached with objectivity, intellectual rigour and a common belief in the rights of all persons involved. Terrorism, particularly that which targets the homeland, challenges many of these assumptions. Its proximity inculcates a sense of fear amongst a population, driving a desire for justice that may manifest itself in demands for retribution or restriction of civil liberties. Responding to the public's real or perceived sense of security requires a careful balance of legislative, investigative and prosecutorial processes that is surgical in its approach but equitable in its application.
This consideration is complicated further by an urgent need to adjudicate what types of politically violent acts constitute "terrorism", particularly in the wake of a resurgence of domestic terrorism activity in the United States (U.S.) . Public opinion plays an important role in how typologies of political violence are classified, with implicit effects on the legislative architecture that facilitates the investigation, prosecution and sentencing of offenders, reflecting both the values and biases of the U.S. as a political community. This study will not explore the origins of or reasons behind disparities in why some offenders are considered terrorists and why some are not. Instead, it will examine closely the outputs of these conditions and how this framing is reflected in criminal justice approaches.
Acts of terrorism were rarely politicised in the U.S. justice system until the events of the 1995 bombing of the Murrah Federal Building in Oklahoma City and the 11 September 2001 (9/11) attacks on the World Trade Center and the Pentagon 1 (Damphousse and Shields, 2007, p. 176). These events became instrumental in revealing how the American public viewed terrorism on American soil and precipitated a variety of responses. While both incidents were reviled by the public, their circumstances reflected in-group, out-group characteristics that may have contributed to divergences in the magnitude of legislative response: the perpetrators of Oklahoma City were white military veterans, while the perpetrators of 9/11 were Arabicspeaking Middle Eastern Muslims. Oklahoma City resulted in more expansive application of existing investigative and criminal statutes, while 9/11 resulted in the establishment of entirely new authorities and parameters for addressing foreign terrorist organisations. One group represented what we will term the Far Right 2 (FR), while the other represented an Islamic Extremist 3 (IE) ideology.
Despite distinctions in belief systems, the two typologies of FR and IE have many overlapping characteristics. Some salient demographic variables are similar (Chermak and Gruenewald, 2015), 4 and both seek a reversion to a prior way of life as opposed to favouring progressive radicalism. There is sometimes congruence on target preference as well, such as many on the FR praising the 9/11 attacks for striking at a perceived "Zionist Occupation Government" (Durham, 2003). Their cultural similarities are as valuable as their differences in terms of evaluating one against the other, and this comparison is a relevant factor to consider as a backdrop to the empirical components of this investigation.
The political environment and its impact on addressing issues of domestic terror Prior to 9/11, terrorism investigations as a whole were less preemptive in nature, and the preponderance of charges fell under conventional criminality statutes. From the 1970s onwards, in the United States, this approach reflected the government's resistance to politicising crime, which stood in contrast to the approaches of many contemporary European and Asian nations. It also reflected the post-Watergate guidelines from Attorney General Edward Levi limiting investigations of American citizens for terrorism and security crimes (Smith et al., 2002, p. 314). A few high-profile domestic terror cases in the 1980s presented opportunities for prosecutors to test a more explicit approach to characterising offenses as political crimes but subsequent acquittals of seditious conspiracy charges in the cases of the United Freedom Front, May 19 th Communist Organization, and white supremacists in Fort Smith, Arkansas discouraged more expansive use of these statutes (ibid., pp. 315-316). Over the course of the 1990s, however, in the wake of the 1993 World Trade Center bombing and other cases, the government learned it could be more effective in securing convictions by characterising internationally affiliated defendants and cases as political crimes. Throughout the rest of the decade, these nuances in prosecutorial approach and outcome in domestic versus internationally affiliated terror cases became more apparent, serving as a prelude to the more explosive divide that these approaches would take once the reaction to 9/11 eliminated many of the investigative, prosecutorial and sentencing barriers that existed in the previous era (ibid.). 5 The 9/11 attack precipitated a fundamental shift in how Americans perceived Islamic Extremism, culminating in unprecedented changes to privacy laws, extensive efforts to counter violent extremism, and a two-decade-long ground war. While significant acts of terrorism often spur reactive political and legislative responses, with resultant tradeoffs between civil liberties and security that can be emblematic of the public's collective emotional state (Heymann, 2016, pp. 428, 435), the depth and longevity of the political and legislative environments emanating from 9/11 far exceeded those arising from any terror event of the previous thirty years. One of the first manifestations of this newly proactive political environment was the implementation of the USA PATRIOT Act, which facilitated a more preemptive approach to countering terrorism affecting the U.S. 6 Outputs of the USA PATRIOT Act expanded the use of electronic searches, roving wiretaps, informants and undercover agents, as well as mechanisms to acquire records. 7 With the statutory imprimatur of preemption, terrorism prevention now shifted to the forefront of the Department of Justice's priorities (Chesney, 2005, p. 21). This has not been without criticism, particularly regarding concerns over privacy, searches, entrapment and government overreach (Norris and Grol-Prokopczyk, 2018;Heymann, 2016;Margulies, 2005). Similarly, the Intelligence Reform and Terrorism Prevention Act (IRTPA) of 2001 addressed legal parameters involving the enforcement of counterterrorism policy. It amended 18 U.S.C. §2332B 8 to include material support as a terrorism offense, broadly expanding the scope of possible application of the terrorism label (McLoughlin, 2010, p. 66). The 2002 Homeland Security Act then established the Department of Homeland Security (DHS) with language that further broadened the authorities of law enforcement agencies (Damphousse and Shields, 2007, p. 191).
The scope of these changes became more apparent as the Islamic State of Iraq and Syria (ISIS) gained regional and international momentum in the 2014-2017 timeframe. Many attacks were planned or carried out by sympathisers within the U.S., 9 and offenders took advantage of rapidly evolving technological and communicative advancements to receive direction and guidance from ISIS leadership in Iraq and Syria. This threat precipitated a greater effort by law enforcement to leverage fully tools allowed by post-9/11 legislation, ranging from the use of intelligence databases to electronic search warrants, to the exploitation of social media (Wray, 2017). Use of these resources enhanced the government's ability to demonstrate more acutely both terrorist intent and connection to a Foreign Terrorist Organization (FTO), which facilitated the application of more severe investigative, prosecutorial and sentencing approaches to such defendants.
An important consideration in analysing the post-9/11 proactive political environment has been the continued and sensationalised reporting of IE attacks in comparison to FR attacks (Powell, 2011). Though FR attacks have continued to occur throughout the United States in the post-9/11 era, they have received far less attention in the media and political discourse (Anti-Defamation League [ADL], 2017, p. 1). Only in the aftermath of the post-2020 election violence has the significance of FR groups become more readily apparent to the general public. Perception of threat is key. In Freilich, Chermak and Simone's survey of American state police agencies, researchers found that while law enforcement identified Islamicaffiliated groups as the highest-ranking perceived threats to state and national security, these groups were not even in the top five most violent groups when measured by level of extremist activity. Conversely, the Far Right Freeman/Sovereign Citizen movement was not one of law enforcement's top five perceived threats but was in the top five when actual extremist activity was compared (2009, p. 463).
While the frequency of FR attacks has fluctuated over the past three decades, the current upward trend in activity warrants particular concern, as it represents a comparable escalation to Islamic-inspired violence in the post-9/11 era (Jones, 2018). The ADL reports that, in 2018, FR attacks resulted in 49 deaths, which represented 98% of domestic extremism killings in the U.S. that year (2019b, p. 14). There have been other years when IE attacks have resulted in a higher percentage of deaths, but the totality of the numbers over the last ten years (2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018) is telling: of 427 domestic extremist-related deaths, 73.3% were perpetrated by FR, 23.4% were conducted by IE, and 3.2% were conducted by Far-Left perpetrators (ibid., p. 16). More recent research, including that by the Center for Strategic and International Studies, also found that despite higher numbers of overall fatalities from IE terrorism in the U.S. during the period of observation , the skew of the 9/11 death toll masks the fact that FR attacks resulted in more deaths per typology in any given year (Jones, Doxsee and Harrington, 2020, p. 3). In 14 out of the 21 observation years, the majority of fatalities resulted from FR attacks, and accounted for 90-100% of fatalities in 11 of those years (ibid.). 10 Despite these statistics, no significant federal legislative changes have been implemented to address FR terrorism. 11 This is further complicated by opposition on both ends of the political spectrum to expand surveillance and investigative authorities that might be used on U.S. Persons. 12 This does not mean that the justice system is entirely neglecting to address FR; as recently as the fall of 2020, the U.S. Federal Bureau of Investigation (FBI) and the U.S. Department of Homeland Security (DHS) acknowledged the rise of the threat of FR subsets, such as white nationalism, with growing consensus that domestic extremism is a comparable or more significant threat to the U.S. than foreign-affiliated terrorism (Wray, 2020;DHS, 2020, pp. 17-19). The National Strategy for Counterterrorism also briefly mentions domestic terrorism as a growing concern (Office of the Director of National Intelligence, 2018, pp. 10, 18). While both typologies, however, present threats to the American people, the "othering" of Islam 13 has resulted in a more politicised environment that elevates the threat of Islamic Extremism more acutely than that of FR terrorism (Jamal, 2008, pp. 114-130).
As demonstrated in the literature, proactive political environments can result from an attack originating from any type of terrorism, but the heightened state of public fear persisting in the post-9/11 environment suggests that public opinion and, thus, government approaches are more reactive to an anti-Islamic political environment than an anti-Far Right one. Thus, the overarching questions this examination seeks to answer are: (A) Does type of terror (Far Right or Islamic Extremist) predict legal response?
(B) Are differences in legal response based on type of terror due to inherent justice system institutional priorities (legal factors, such as charged offense severity or investigative resources applied) or other factors (extralegal, such as country of origin)?
Answering these questions requires an understanding of the legal architecture that shapes the government's jurisdictional limits, as well as an appreciation for the theoretical frameworks under which we can evaluate legal responses to terrorism.

Legal background
Jurisdiction. Terrorists in the United States can be prosecuted under federal law, state law or both. State laws regarding terrorism vary considerably, and the statutory requirements for proving terrorist intent differ by state. Generally, at the time of indictment, state and federal authorities collectively decide the jurisdiction(s) in which the suspect will be charged. If the government judges state resources and legal avenues to be sufficient in gaining a conviction, the suspect will likely face state charges (Donoghue and Kayyem, 2002). If federal assets have been leveraged over the course of an investigation, there is a higher probability that the suspect will be charged and tried in federal court; such investigations often transcend state lines, thereby depriving state prosecutors of complete geographic jurisdiction, known as venue. Another consideration is the relative admissibility of evidence. For example, investigative techniques may be legal under federal law (e.g. wiretapping or other electronic surveillance) but inadmissible in states with more restrictive laws governing such techniques.
If an act meets a threshold of national security, foreign affairs or domestic defence, it may also more appropriately fall under federal jurisdiction (ibid., p. 2). This delineation, particularly as it relates to terrorism statutes, results in a high number of domestic right-wing cases being tried in state court and a large percentage of IE cases being tried in federal court, although this is far from an exclusive delineation. This study will illuminate some of these trends at the federal level, which provides the most consistent metrics for evaluation.
Investigation. The FBI is the lead federal agency responsible for the investigation of terrorism cases involving U.S. Persons (USP). 14 The FBI standard for determining if a case meets the threshold for a terrorism investigation is the definition articulated in the Code of Federal Regulations (28 CFR §0.85), 15 which describes terrorism as 'the unlawful use of force or violence against persons or property to intimidate or coerce a government, the civilian population, or any segment thereof, in furtherance of political or social objectives' (DOJ/ FBI, Terrorism 2002. This study will apply this definition exclusively, as it is the U.S. government threshold that a case must meet in order to be included in the dataset used.
Meeting the de facto definition of terrorism to qualify as a terrorism investigation, however, does not necessarily mean that a defendant will be charged with a crime of terrorism, due to the challenges of proving ideological intent beyond a reasonable doubt.
Prosecution. Turk introduced two primary prosecutorial approaches for terrorism cases: explicit politicality, which identifies the defendant as a terror suspect through the charges levied and the ensuing media portrayal, and exceptional vagueness, which is a strategy of conventional criminality (1982). Application of each strategy indicates the prosecution's assessment of the public's willingness to entertain discussions of motive and intent. Studies have shown that, historically, prosecutors revert to a conventional criminality approach in the aftermath of a major terrorist event, and that defendants behave more like traditional offenders in these cases (Damphousse and Shields, 2007). Studies have also shown that application of an explicit politicality prosecution strategy has been historically more likely to be employed with international as opposed to domestic terror defendants, particularly in the pre-9/11 era of 1980-1998 (Smith et al., 2002, p. 324). This is important to consider for our study since IE defendants are often treated like international terrorists because of potential linkages to FTOs.
Charging. Much of a prosecution's strategy can be discerned from the charges levied, as the two are closely tied. Only those defendants who have a connection to an FTO can be charged with federal terrorism crimes (ADL, 2017, p. 8). This distinction used to be more apparent, as geography, communicative ability, and the investigative capability to establish an international nexus were more limiting in the past. The Antiterrorism and Effective Death Penalty Act started to blur the lines between distinguishing domestic from international terrorists when it was implemented in 1996 (Smith et al., 2011, pp. 6-7). Since then, technology has facilitated a rapid rise in communicative and material support capability in terror groups 16 that decidedly transcends international borders, representing opportunity not only for homegrown extremists to reach out to FTOs, but for the U.S. government to intercept and prove international nexus on a greater scale. The alternative to terrorism or Hate Crime charges (which also require proving ideological intent and are also eligible for sentencing enhancements) remains conventional criminal statutes. Prosecutors often assess this approach as having a greater certainty of conviction, charging defendants with crimes such as firearms violations, possession of explosives or financial crimes (Bradley-Engen et al., 2009, p. 436). Juries are often responsive to cases that exhibit tangible proof, such as recovered weapons (Reskin and Visher, 1986, p. 429), and access to this type of evidence often shapes prosecutorial approaches in many types of federal cases (Reskin and Visher, 1986). Pursuing a conventional criminality approach also eliminates the potential that a jury might be sympathetic to an alleged terrorist's cause, while reserving the opportunity to introduce ideological intent in the non-jury sentencing phase.
Sentencing. The sentencing of terrorists has many nuances and controversial aspects and is well nested within the broader debate about civil liberties versus national security (Brown, 2014, p. 545;Ahmed, 2017). Sentencing policy reflects public sentiment, as it relates to counterterrorism efforts, especially since the U.S. Sentencing Commission has significant influence from elected officials. In 1984, Congress passed the Sentencing Reform Act, which sought to reduce disparities in sentencing that arose from judicial discretion. The matrixed assignment of offense types and levels evolved into the United States Sentencing Guidelines (USSG) in 1987.
The mandatory nature of the Guidelines prompted repeated legal challenges over the ensuing years, complicated by the addition of the 'terrorism enhancement' to USSG §3A1.4, which mandated an increase to the base level offense by 12 levels, or to at least Level 32, and a criminal history category of VI (USSG §3A1.4), 17 requiring a sentence range of 210 to 262 months. 18 Since exclusively domestic terrorists could not and still cannot be charged with terrorism under federal criminal statutes, the imposition of the enhancement required demonstrating to the sentencing judge that the defendant's actions were 'calculated to influence or affect the conduct of government by intimidation or coercion, or to retaliate against government conduct' (18 U.S.C. §2332b(g)(5)(A)). The sentencing phase has a decreased standard of proof (Bradley-Engen et al., 2009, p. 436), and must only establish intent by a preponderance of evidence -i.e. more likely than not -rather than by the exacting trial standard of beyond reasonable doubt. While there has been litigation challenging this process (Brown, 2014, pp. 529-533), the USSG are now generally accepted as advisory and employed as such. 19 Structural-Contextual theory as an evaluative framework Legal approaches to countering, preventing or deterring terrorism are direct manifestations of public opinion and public policy. There are many mechanisms by which governments can choose to address the perceived threat, from a militaristic approach to one that treats terrorism as an element of criminality. 20 The U.S. justice system finds itself somewhere in the middle, implementing legislative solutions that seek to balance fundamental security interests against well-established civil rights and procedural protections.
Structural-Contextual theory is one framework that considers the effect of proactive political environments on sentencing outcomes. Hagan defines a proactive political climate to be one in which 'the surrounding political environment has mandated departures from normal criminal justice operations', and involves 'the imposition of political power, sometimes targeting the prosecution of a particular form of crime and criminal ' (1989a, p. 130).
Hagan first introduced Structural-Contextual theory as an attempt to reconcile elements of Consensus theory and Conflict theory, which he maintained had little empirical support as independent theories (1989a, 1989b). Consensus theory suggests that legal variables (crime severity, prior record etc.) are more powerful in predicting criminal justice outcomes, 'reflecting the influence of broadly shared societal values in the punishment of criminal norm violations' (ibid., p. 116). Conversely, Conflict theory suggests that extralegal variables (demographic characteristics) are more influential in affecting these outcomes, 'reflecting the influence of power imbalances in the punishment of crimes that posed threats to existing power relationships' (ibid.). Hagan suggested that explained variance in sub-systems of the justice system increases when components of the government work more interdependently, representing a tightening effect that sees investigative and prosecutorial elements working closely together to achieve a conviction (ibid., p. 129).
Subsequent studies have evaluated Structural-Contextual theory in the context of terrorism. Smith and Damphousse applied it in 1998 to compare terrorists to non-terrorists, situating terrorism cases within the proactive political environment described by Hagan. Smith and Damphousse concluded that the framework was generally applicable to understanding sentencing outcomes for terrorists compared to non-terrorists. 21 Other researchers built upon this baseline. Shields et al. (2006) used Structural-Contextual theory to evaluate guilty plea rates amongst terrorists and non-terrorists. They found that terrorists were convicted at trial at twice the rate of comparable non-terrorist offenders, and six times the rate for federal defendants overall. In the context of previous research about the circumstances under which prosecutors offer plea deals, Shields et al. assessed that the higher rates of conviction at trial were indicative of the proactive political environment and severity of the crimes committed (ibid., p. 272). More recently, Structural-Contextual theory was evaluated by Murray in her 2016 examination of domestic terror conviction rates, where she found significant differences in between-group conviction outcomes amongst domestic terrorists. She documented that ecoterrorists 22 received more lenient outcomes than left-and right-wing terrorists, which she partially attributed to the increased media attention and subsequent coupling of the justice system that left-and right-wing terrorism generally receive (ibid., p. 85).
Evaluating between-group variables of domestic terrorism. There is a gap in analysing outcomes amongst terror groups primarily operating in a domestic setting (Murray, 2016, p. 77), particularly between FR and IE. Previous studies have compared in-group demographic variables (Chermak and Gruenewald, 2015) or have focused on right-wing versus left-wing typologies (Smith, 1994;Handler, 1990). Comparative approaches that have studied the spectrum of response to FR and Islamist-affiliated terrorism in the United States have focused on specific elements of the justice system. For example, Norris and Grol-Prokopczyk's study specifically examines instances of entrapment 23 from 1989 to 2014, finding that despite a doubling in the number of sting operations after the Oklahoma City bombing, entrapment indicators dropped, whereas the rise in sting operations after 9/11 featured high numbers of entrapment indicators (2018, p. 243). Murray's study (2016) compares conviction outcomes within the domestic terrorist population, finding significant differences amongst left-wing, right-wing and ecoterrorist defendants, but does not evaluate IE. Bradley-Engen et al. (2009) examine disparities between terrorists and non-terrorists prior to the revised USSG, which has implications for this study because right-wing extremists often "look" like non-terrorists in how they are prosecuted, while IE are more likely to be charged as terrorists. Rondon (2018) examines mental health defences in right-wing versus Islamist-affiliated terrorism cases and finds that a mental health defence is much more likely to be accepted with a non-Arab defendant.
Additional studies applicable to this examination did not compare FR and IE prosecutions, but rather domestic versus international cases. As most domestic IE defendants can now be linked to FTOs through the use of technology, they are often treated in the courts as international terrorists. 24 Similarly, studies have looked at sentencing outcomes, concluding that international terrorists are more likely to be sentenced to 20+ years in prison compared to domestic terrorists (Smith et al., 2002, p. 329). Murray notes that these differences raise 'further questions about how the origin of terrorism and nationality of the defendant sway court decisions in terrorism-related cases' (2016, p. 77). Specifically examining the conviction outcomes of cases that have arisen from the USA PATRIOT Act, Tauber and Banks (2015) find that judges are significantly more likely to defer to the government's case when the terror threat appears to be heightened, illustrating an implicit reflection of the political environment. These studies provide a limited but helpful basis for comparison.
Given this baseline, the present study fills a critical void in the literature with its specific evaluation of FR versus IE legal outcomes in the U.S., as two typologies of terror existing in parallel U.S. political environments. Moreover, it offers an empirical assessment to the field of terrorism studies, the lack of which is often identified as a limitation (Lum et al., 2006, pp. 491-492), and adopts the lens of a criminological approach to the study of terrorism, which is also under-represented (Freilich and LaFree, 2015, p. 1).

Observable themes and applicability of Structural-Contextual theory to this study.
While not an exclusive component of this study, Structural-Contextual theory is presented as a potentially informing and useful framework. In order to evaluate the applicability of this theory and the concurrent questions that this project poses, there are general themes under which we will evaluate the results: (1) Prosecutorial approach. This study will glean insight into the prosecutorial approaches described above through an examination of the charges levied against each type of defendant.
Similarly, it will evaluate success rates for leveraging each type of approach. The researcher will employ the descriptors of explicit politicality or conventional criminality.
(2) Plea and conviction outcomes. Previous studies have evaluated plea rates amongst terrorists versus non-terrorists, 25 as well as domestic versus international terrorists, 26 and have interpreted these results as evidence of tightening or coupling within a proactive justice system. This examination will address findings in the context of how plea and conviction rates may reflect this coupling.
(3) Sentencing outcomes. Previous studies have evaluated the sentencing of terrorists versus non-terrorists, and specifically examined how explained variance in sentencing outcomes can be elucidated through Structural-Contextual theory. Bradley-Engen et al. (2009) note that terrorists receive longer sentences than non-terrorists (p. 434), while Smith and Damphousse find punishments to be much more severe for terrorists than non-terrorists (1996, p. 312). Smith and Damphousse conclude that 'the official label of "terrorist" is not only a significant predictor of sentence length but emerges as the dominant explanatory variable in the analysis' (ibid., p. 289). This study will determine if significant differences in sentencing outcomes exist between FR and IE indictees, and to what level that variance can be explained.
(4) Reflections of a proactive political environment. Building on Hagan and Smith's initial descriptors, Bradley-Engen et al. (2009) provide an illuminating history of how major attacks in the U.S. have resulted in proactive political environments. They describe that terrorism investigations are not "general" criminal justice operations, and that such investigations focus political power on the prosecution of a specific type of offender -key characteristics of a "proactive political environment". When an act is officially labelled as "terrorism" by the government, proactive justice procedures are set in motion' (ibid., p. 438). This study will identify reflections of a proactive political environment by evaluating factors such as magnitude of government resources applied to a case, which can be indicative of preemptive investigative processes and coupling of the justice system. 27

Methodology
Quantifiable research sub-questions. The overarching research questions evaluate whether the type of terror predicts the government's legal response, and what factors contribute to any differences. In order to assess this broad construct, we ask the following quantifiable questions: (1) What type of charges (categories of crime) are most commonly used to prosecute Far Right versus Islamist-affiliated terrorists?
(2) How does type of terror predict overall case outcome?
(3) How does type of terror predict the outcome of the highest-charged criminal count?
(4) How does type of terror predict sentencing length?
(5) Does type of terror predict magnitude of legal resources applied by the government?
(6) Is the relationship between type of terror and sentencing length due to the magnitude of legal resources?
(7) Is the relationship between type of terror and government success of conviction due to the magnitude of legal resources? (8) To what extent do legal and extralegal variables explain variance in sentencing for each type of terror?
Hypothesis. The researcher predicted that there would be significant differences in approaches and outcomes to prosecuting and sentencing FR versus IE, with more severe outcomes expected for IE. This hypothesis was based on the expressly proactive political environment evident in post-9/11 legislation and legal architecture. Given this proactive political environment, the researcher expected to find support for Structural-Contextual theory in explaining sentencing outcomes.
Dataset. The primary data source for this project was coded data from the American Terrorism Study (ATS) at the University of Arkansas Terrorism Research Center (TRC). The ATS maintains a comprehensive database of federal terrorism cases filed in the United States since 1980 and is the longest running research project on terrorism in the country. Since its inception, ATS has been coordinated with or sponsored by the FBI, the U.S. Senate Judiciary Committee, the National Institute of Justice and DHS, resulting in the comprehensive inclusion of all officially designated terrorism cases through to 2004. Since 2004, the method of identifying and including cases has involved evaluation of press releases from U.S. Attorney's Offices around the country for cases that continue to meet the FBI definition of terrorism. Once these cases are identified, the ATS compiles indictments, criminal complaints, sentencing reports and other official court records. Data is extracted manually by reading these court records, after which information is coded into an Oracle 11g relational database at the court case, indictee and count level. Each case goes through three levels of review to ensure multi-rater reliability in coding.
This research used court records as the primary source of data because they have been validated in the literature as a reliable data source for evaluating these types of trends (Freilich et al., 2014, pp. 374, 383;Murray, 2016, p. 79). As case outcome may or may not be indicative of outcome for most severe count (i.e. case may have resulted in conviction at a lower count), this examination distinguished between outcomes for highest count versus the overall case.
The population for this study included data for indictees classified as FR and IE for the period from 12 September 2001 to 7 February 2019, which represented the totality of the post-9/11 period at the commencement of the study. The ATS classifies cases as FR when they exhibit qualities of being 'fiercely nationalistic', 'reverent of individual liberty', hyper-vigilant of threats to national sovereignty or personal liberty, and militarily prepared for an attack. In this dataset, FR groups included but were not limited to the Army of God, Aryan Brotherhood, Aryan Nation, Creativity Movement, Freemen, Hutaree, Ku Klux Klan, Sovereign Citizens and White Aryan Resistance (WAR). FR cases were all generally characterised by the FBI as domestic terrorism cases.
The ATS classifies cases as IE when they exhibit qualities of being singularly accepting of the Islamic faith, focused on the need to wage jihad, believing that Islam is under attack, and maintaining that the West is responsible for the exploitative corruption, oppression and humiliation of Islam across the world. 29 In this dataset, IE groups included but were not limited to Al-Qaeda, Al-Shabaab, Al-Ummah, Boko Haram, Hamas, Hezbollah, Islamic State of Iraq and Syria (ISIS), Lashkar-e-Taiba (LeT) and the Taliban. For both FR and IE, not all indictees were formally affiliated with terrorist groups but their individual actions mirrored group objectives. 30 Where data was available, approximately 85-90% of cases involved USP defendants, indicating that the context of their crimes could be classified as domestic in most circumstances, despite the trend of charging these indictees as affiliates of FTOs. This study did not include any terrorists falling under the jurisdiction of the U.S. military legal process, and only accounts for cases where the U.S. justice system brought charges that could be prosecuted in a U.S. civilian court of law.
Each indictee was included as a separate occurrence, since some cases had multiple indictees with varying charges and outcomes, and each represented one individual person per individual case. An individual person only occurred more than once in the data if they were indicted in multiple, separate cases, since each case represented a distinct legal approach and process. This resulted in 487 FR and 630 IE indictees, in a total of 1,117 defendants. The terms "indictee" and "defendant" are used interchangeably in this examination.
Due to missing data for some variables, the population size for certain analyses only included those cases with results. The research limited analysis to one charge per indictee, i.e. the highest severity count. This was to eliminate replication, so that the data was not skewed by indictees appearing multiple times in the database for multiple counts within the same case. This method of using only the highest count has been employed in similar research using data from TRC/ATS (Murray, 2016, p. 80). While this strategy is limiting in the sense that it does not provide the full scope of the type of charges against an indictee, limitations were mitigated by accounting for number of counts in some analyses, and by annotating when a defendant pleaded to a count lower than the most severe count charged.
Variables. The variables considered in this analysis were: type of terror (FR or IE), case outcome, highest count charged, category of crime (re-categorised from chapter in U.S. Code, see Appendix C), highest count outcome, highest count severity, total sentencing length, case length, use of informant, use of undercover agent, number of counts, magnitude of resources, gender and citizenship status. All of these were considered legal variables except for gender, citizenship status and type of terror, which were considered extralegal.
Analysis. The researcher used chi-square, logistic regression, multiple regression, independent samples t-tests, mediation models and descriptive statistics. Mediation models used the PROCESS macro in SPSS. For most questions, analysis was performed using two types of tests, allowing to determine both association and prediction/directionality. Results were considered statistically significant if p<0.05, meaning that there was a greater than 95% likelihood that they were not due to chance alone.

Results
The sub-questions of this research project are quantifiable and examined individually here. These individual results are then incorporated into the research study's collective themes that will be synthesised later in Discussion.
(1) What type of charges (categories of crime) are most commonly used to prosecute Far Right versus Islamic Extremist terrorists? This sub-question was designed to illustrate the general categories of charges that each type of terrorism most commonly faces in order to contribute to an understanding of the prosecutorial approach. Its findings give context to subsequent sub-questions that evaluate the efficacy and output of the prosecutorial approach, such as conviction rate and sentencing length. Each indictee was evaluated using the highest charge for the case, since the lead count in an indictment is the most severe (see Table 1; see Figure 1).
Among IE cases, ideological crimes were the most common type of charge, with 45.2% of IE defendants receiving this as their most severe charge; conversely, only 9.4% of FR defendants were charged with an ideological crime as their highest charge. Among FR cases, contraband charges were the most common, with 31.8% receiving this as their most severe charge, whereas only 8.9% of IE had this as their highest charge.
Overall, the results illustrate a difference in prosecutorial approach between groups. We can consider ideological crimes to represent an explicit politicality approach, and all other categories to represent a conventional criminality approach. Here, 90.8% of FR terrorists were charged under a conventional criminality approach (no requirement to prove ideological intent), compared to only 55.2% of IE.
Heavy leverage of ideological/terrorism statutes against IE may be interpreted as being partially representative of the post-9/11 proactive political environment, demonstrating the government's assessment that the chance of conviction for a political crime is much higher for IE. Similarly, the prosecution may believe that it has the evidence to convict successfully on counts that require proving ideological intent, and the evidence is often the result of tightly coupled investigative and prosecutorial processes. Meanwhile, the fact that most FR cases are headlined by a contraband charge may demonstrate that the prosecution does not believe that (1) it has accrued the evidence to prove ideological intent, or (2) it can convince a jury or judge of ideological intent, regardless of evidence. It may also be indicative of the fact that some of these FR cases start within non-terrorism investigatory lanes, such as with the Bureau of Alcohol, Tobacco, and Firearms or the Drug Enforcement Administration, which could demonstrate (1) more surety of conviction with tactile contraband evidence, i.e. the hefty sentence likely stemming from the number of weapons, weight of drugs or other quantifiable contraband obviates the need to establish intent, (2) that the defendants only came to the attention of law enforcement because of conventional criminal behaviour, as opposed to expressions of ideological intent, or (3) that IE may be more likely to employ nontraditional weapons, such as vehicles.
(2) How does type of terror predict overall case outcome? Case outcome is a valuable metric by which to assess government success against terror suspects. Conviction demonstrates that the government successfully constructed a case with convincing evidence and an appropriate prosecutorial approach. Conviction can take the form of either a plea deal or a trial conviction. Plea deals are advantageous for the government because they reduce the amount of time and resources that must go into a trial. Plea deals are also advantageous for defendants because they generally result in less severe sentences (Smith, 2011, p. 23;Rodgers, 2019;Johnson, 2003, pp. 456-457), and defendants often accept such deals if the evidence is overwhelming or if they do not believe they can win at trial. A plea deal, however, is a two-way agreement, and the government may not offer one to a terrorist if they want to make a deterrent statement, or if they are willing to expend the resources and attempt to achieve a more severe sentence at trial. Similarly, many terrorists may not accept an offered plea deal if they want to use the trial to showcase their ideological causes.
Trial conviction illustrates a successful attempt by the government to make a case to the finder of fact, typically a jury of peers, 32 that a crime was committed beyond reasonable doubt, and requires the dedication of significant time and resources. On the opposite end, acquittal generally represents government failure, while dismissal generally signals an assessment by the government that the case no longer has the strength to make it worth pursuing prosecution.
In this dataset, cases that have not been adjudicated are not included in this or any subsequent analysis. This question only evaluates final case outcome, as opposed to outcome of the highest charged count (see Question 3 for results of highest charged count; see Table 2 for each type of terror's distribution among possible outcomes).
Between-group differences within guilty plea and conviction rates foreground the following comparative analysis: Plea deals. The terrorist plea rate has fluctuated over the past 30 years, often in response to changes in the USSG, but has consistently been below the national average of 97% amongst all federal criminal cases (United States Sentencing Commission, 2018). The results of this analysis show that the plea rates for both FR (56.5%) and IE (73.6%) are both well below the national average. When we evaluate these rates in comparison to one another, chi-square analysis indicates that types of terror significantly differ in whether indictees plead guilty (see Table 2). Linear regression further indicates that type of terror significantly predicts a higher occurrence of guilty pleas for IE (see Table 3). This may indicate that (1) the volume of evidence against an IE defendant may be more convincing, or (2) that an IE defendant does not assess that he or she will win the sympathies of a judge or jury at trial. Since we cannot determine how many plea deals within each group were offered versus how many were accepted, 33 it is difficult to consider if IE and FR are offered plea deals at different rates, or if one group felt stronger about accepting/rejecting plea deals.
Trial convictions. Looking at the data in isolation, trial conviction rates reflect higher rates for FR (34.0% of FR cases versus 20.3% of IE resulted in trial conviction). Evaluating this metric alone, however, would be misleading, since most of the cases that would have gone to trial with a high likelihood of conviction were likely to have already been adjudicated with a plea deal. In this examination, trial conviction rates alone were not evaluated because of this potential for misrepresentation. It can be helpful to compare trial conviction rates against plea deals for other factors though, such as in evaluating sentencing outcomes (see Question 4).
Overall government success rate. When we combine the results of plea deals plus trial convictions, we can see the totality of a "behind bars" rate, representing a successful attempt by the government to incarcerate or generate punitive measures against terrorist defendants. When plea deal and trial conviction rates are added within each group, they appear nearly equal (FR: 56.5% plea deals plus 34.0% trial convictions equals 90.5% of FR cases; IE: 73.6% plea deals plus 20.3% trial convictions equals 93.9% of IE cases). Chi-square analysis, however, shows that types of terror actually differ significantly in whether the government's attempt at conviction is successful, with IE defendants being significantly more likely to end up with a conviction (see Table 2). Linear regression confirms that being an IE indictee significantly predicts an overall conviction outcome (see Table 3).
Even though IE is more predictive of conviction than FR, the cumulative results are noteworthy in that they demonstrate rates of government success against both groups above a rate of 90%; prosecutors must overcome a large delta after starting with a much lower FR plea rate, and this final result likely represents a commitment by the government to expend the resources during the trial phase to achieve a conviction no matter what the type of terror. Factors that could have affected these results include the fact that many domestic terror cases are tried at the state level rather than the federal. This skews the population that can be evaluated, since FR cases that reach the federal level tend to be stronger. IE cases, meanwhile, more often than not, find themselves at the federal level because expanded investigative and statutory allowances are derived from federal law. There are also opportunities to leverage federal intelligence capabilities, which may not be applicable to FR groups with no apparent international ties.
(3) How does type of terror predict the outcome of the highest charged criminal count? The intent of this question was to determine prosecutorial success rate with the highest charged crime, as opposed to looking at the overall case. This gives insight into the success rate of choosing to charge defendants under a certain category of crime, building on the data from Question 1 and elucidating the efficacy of the prosecutorial approach. Guilty pleas and trial convictions are interpreted as successes for the state, guilty pleas on lower counts are interpreted as partial successes (and may or may not indicate a failure of the prosecutorial approach, depending on the lower count to which the defendant pleads), while acquittals/ dismissals are interpreted as unsuccessful. Descriptive statistics illustrate differences in highest count outcome within each typology (see Table 4).
Chi-square analysis reveals that types of terror differ significantly in whether or not indictees plead guilty to the highest count with which they are indicted (see Table 4). Linear regression further indicates that type of terror significantly predicts a higher occurrence of guilty pleas to the highest count for IE (see Table 5). After controlling for count severity in a multiple regression, the results indicate that type of terror still significantly predicts whether indictees plead guilty to the highest count with which they are indicted, with IE significantly more likely to plead to the highest count (see Table 5).
When considering the cumulative percentage of plea deals plus convictions for highest count, 34 chi-square analysis reveals that types of terror do not significantly differ in whether or not indictees plead guilty or are convicted at the highest count with which they are indicted (see Table 4). Linear regression validates that type of terror does not significantly predict whether indictees plead guilty or are convicted at the highest count with which they are indicted (see Table 5). After controlling for count severity in a multiple regression, the results indicate that type of terror is still not a significant predictor of highest count conviction outcome (see Table  5). These results contrast with the results for overall case outcome, where the government's success rates against FR and IE cases are significantly different. This could indicate that while prosecutors are almost equally successful with their primary approach (represented by highest count), FR cases become weaker once the primary approach does not work, reducing comparative efficacy in the final outcome.
With these results in mind, it is helpful to understand government success rate for achieving plea or conviction on the highest count within each category of offence. This can tell us why or how a certain approach for a certain typology of terror may or may not be effective. When broken down into each category, the sample sizes are underpowered to determine statistical significance. Results can be interpreted as descriptive statistics (see Table 6).
The government's overall success rate (plea or trial conviction), when using an explicit politicality approach, is higher within the IE group (91.8% of those charged with ideological crimes) than the FR one (73.8% of those charged with ideological crimes). For conventional crimes, the results vary. While there is a higher success rate of achieving a plea or trial conviction for IE within violent conventional crimes (78.9% of IE versus 64.3% of FR), other categories are remarkably similar in outcome, such as contraband (81.5% of FR and 77.8% of IE). Notably, IE defendants also plead guilty at a higher rate than FR ones in every category except for international travel or status violations, which has a population size of 1 and a plea rate of 100%. Even when IE defendants plead at a higher rate, FR defendants end up with a higher rate of incarceration within categories such as contraband, financial crimes and other general crimes.
(4) How does type of terror predict sentencing length? Sentencing length is revealing because it represents the government's assessment of the severity of a defendant's crime and the danger he or she may pose to the public. Descriptive statistics provide insight into the scope of sentencing length. After adjusting sentences for Life, Natural Life and Death to be just beyond the longest ordinal sentence in months, the disparities between groups become readily apparent (see Table 7).
The median overall sentence length for IE is double the median overall sentence length for FR (96 versus 48 months). Additionally, while the median sentence length for FR defendants convicted at trial is double the median sentence length of FR defendants who plead guilty (72 versus 36.5 months), the median sentence length for IE convicted at trial is nearly five times the median sentence length of IE defendants who plead guilty (288 versus 60 months). This variation may help explain why IE defendants are more likely to plead guilty.
In order to confirm whether these results contained a predictive quality, the researcher conducted multiple regression analysis, controlling for count severity, number of counts, status of U.S.-born citizenship, and whether an indictee pleaded guilty or was convicted at trial. The results show that being an IE indictee significantly predicts a longer sentence length, even after accounting for these other variables (see Table 7). In their totality, the results of this analysis raise further questions about the magnitude of terrorism enhancement in influencing sentencing guideline recommendations: even though terrorism considerations can be introduced during the sentencing phase, distinct from whether or not a defendant was charged with a crime of terrorism (a large percentage of IE defendants), these results show a likely disparity in whether the terrorism label is applied equitably to both groups during the sentencing phase.
(5) Does type of terror predict magnitude of legal resources applied by the government? The government's priorities can often be found in where it spends its resources. The intent of this question was to determine if each type of terror received a different level of resourcing in the investigation and prosecution phases. Resources account for time, money and risk. As described in Methodology, magnitude of legal resources was determined by assigning values to the use of informants and undercover agents, case complexity (over/under mean case length in days) and number of counts (over/under mean number of counts). Each case received a cumulative value of 0-4 based on these characteristics. Percentages of each value within FR versus IE revealed definitive differences in magnitude of resources applied (see Table 8).
Analysis via an independent samples t-test indicates that type of terror significantly predicts magnitude of resources used (t=-9.724, df =1006.403, p<0.001). 35 The researcher used multiple regression to evaluate prediction and directionality, while controlling for U.S.-born citizenship status. This analysis reveals that type of terror significantly predicts magnitude of government resources applied, even when U.S. citizenship status at birth is accounted for: IE indictees are significantly more likely to have more resources applied than FR indictees. 36 This result gives additional context to manifestations of a proactive political environment around combatting Islamic Extremism, indicating that executive direction, appropriation of funds, and enabling legislation that facilitates more expansive investigative authorities directly translate into more effort expended on this typology of terror.
(6) Is the relationship between type of terror and sentencing length due to the magnitude of legal resources? The researcher utilised a mediation model to analyse if the relationship between type of terror and sentencing length was in fact due to the magnitude of legal resources, controlling for U.S. citizenship at birth, count severity, and whether the indictee pleaded guilty or was convicted at trial. After accounting for these controls, the magnitude of legal resources used partially mediates the relationship between type of terror and sentencing length (see Figure  2). This tells us that while legal resources do not explain entirely why IE defendants receive significantly longer sentences, the use of a greater magnitude of legal resources for IE indictees plays a significantly large explanatory role as to why IE terror results in significantly longer sentencing periods.
(7) Is the relationship between type of terror and government success rate due to the magnitude of legal resources? After accounting for U.S. citizenship at birth and count severity, magnitude of legal resources partially mediates the relationship between type of terror and plea/conviction versus acquittal/ dismissal outcome (see Figure 3). This means that while legal resources do not explain entirely whether the government is successful in overall conviction (guilty plea or trial conviction), the use of a greater magnitude of legal resources for IE indictees plays a significantly large explanatory role as to why IE terror results in significantly greater conviction rates than FR cases.

(8) To what extent do legal and extralegal variables explain variance in sentencing
for each type of terror? Structural-Contextual theory is predicated on an understanding of how various factors influence explained variance in sentencing. In order to evaluate explained between-group variance in sentencing outcomes, the researcher assessed each type of terror independently against legal and extralegal variables. The dataset for each consisted of those cases that resulted in a plea or trial conviction. Legal variables included count severity, plea versus trial conviction, and magnitude of resources. Number of counts was incorporated as a component of the magnitude of resources variable. Extralegal variables included U.S.-born citizenship and gender.
Regarding legal variables, there are differences in significance based on type of terror. For FR cases, magnitude of resources does not account for a significant degree of the variance in sentencing outcomes, after accounting for count severity and plea or conviction. Count severity alone explains 5.4% of the variance in sentencing outcomes; 37 plea versus conviction explains 6.7% of variance while accounting for count severity; 38 and magnitude of resources explains 6.6% of variance while accounting for count severity and plea versus conviction. 39 In contrast, legal variables are significant in predicting IE sentencing outcomes. Count severity alone explains 8.0% of the variance in sentencing outcomes; 40 plea versus conviction explains 22.3% of variance while accounting for count severity; 41 and magnitude of resources explains 23.8% of variance while accounting for count severity and plea versus conviction. 42 The difference in explained variance in sentencing outcomes between FR and IE is stark, and gives additional insight into reflections on a proactive political environment: Structural-Contextual theory posits that as the political environment increases its focus on a typology, explained variance also increases.
Extralegal variables reflect no significance for either type of terror in this analysis. For FR cases, U.S. citizenship at birth explains -0.1% of variance, and gender explains -0.4% after accounting for U.S. citizenship. For IE cases, U.S. citizenship at birth explains 0.1% of variance, and gender explains 0.4% after accounting for U.S. citizenship. The fact that these results illuminate differences in legal or extralegal explanatory power adds to our understanding of how different factors affect variance.

Discussion
The collective results of this examination demonstrate that there are clear differences in prosecutorial approaches for FR versus IE cases. Overall, IE defendants are 4.8 times more likely to be charged with offences that require demonstration of ideological intent, such as charges of terrorism. Prosecutors may make assessments of approach based on recent historical success rates of charging defendants under this explicit politicality approach, as opposed to conventional criminality: the higher success rate of ideological charges against IE defendants over the past 18 years is one such indication. This data does not offer conclusive results on why the success rate is higher, but possible explanations include preemptive investigative techniques and resource allocation directed at IE defendants, which may result in more substantive evidence, or prosecutor assessment that the likelihood of jury conviction for an ideological crime is higher for IE. The disparity between IE and FR also mimics previous studies on international and domestic terrorists; here, IE cases result in similar outcomes to international cases, even if they took place entirely in the United States. The designation international assumes different connotations in an era where technology facilitates transnational relationships, allowing the U.S. government to characterise a terrorist on U.S. soil under the same parameters as one who has never stepped foot in the United States. This is interesting in light of a growing recognition that some FR movements, such as the White Power movement, have transnational affiliations (Belew, 2019;ADL, 2019a), which are facilitated by recent innovations in technology and communication across borders. Still, perhaps it is not expedient for prosecutors to delineate these affiliations in court when the international connection is not designated by the U.S. Government as an FTO. As of January 2021, it appears that the government has designated only one White Power organisation, the Russian Imperial Movement, as an FTO (U.S. Department of State, 2020). Nonetheless, this dichotomy reveals much about the U.S. as a political community that is so ready to apply tangential applications of "foreign" involvement to IE groups, while FR affiliations that would open up comparable levels of investigative, prosecutorial and sentencing authorities go largely unexplored.
Damphousse and Shields' study concluded that the more terrorists are treated with a conventional criminality approach, the more likely they are to plead guilty, consistent with rates of non-terrorists (2007, p. 191). Our data suggests the opposite, as the group with the significantly higher rate of conventional criminality charges (FR) also has a significantly lower rate of plea deals (56.5%, compared to 73.6% for IE). In this case, it is IE defendants who exhibit plea rates more closely aligned with non-terrorists.
Crime severity and prior record were intended to be equalising variables in the USSG. Our data demonstrates that many other factors contribute to sentencing length as well, cutting across a period of time when sentencing guidelines were either mandatory or advisory. Despite the fact that crime severity is viewed as one of the strongest moderating variables in evaluating sentence length, Smith and Damphousse (1998, p. 73) suggest that it may be less important than the role of a proactive political environment in predicting outcome. Our research is consistent with this assessment, as crime severity is consistently used as a control but results still reflect between-group differences more attributable to the nature of the political environment. Additionally, it is noteworthy that the group with a higher rate of plea deals (IE) still has significantly longer sentencing (median 96 months for IE, compared to 48 months for FR), when plea deals generally result in more lenient sentences because of the defendant's acceptance of responsibility and willingness to avoid putting the government through trial (Smith, 2011, p. 23;Rodgers, 2019).
We began this examination with an assumption that IE cases were tried under the construct of a proactive political environment, and this study illuminates many manifestations of such an environment. The significantly higher magnitude of resources applied to IE cases is one example of the mantra of preemption that characterises this environment. Notably, our finding that IE defendants have more resources applied in a preemptive manner, yet still receive significantly longer sentence lengths, complicates the findings of Jackson (2011), who concluded that preemptive investigative approaches actually result in lesser sentences because of early interdiction and less evidentiary strength.
In partial support of the Consensus theory, legal factors explain many variations of case or count outcome in this study. Count severity and magnitude of resources, for example, significantly affect outcomes. In marginal support of the Conflict theory, the extralegal factor of type of terror has similar effects, while other extralegal factors do not influence outcomes. The researcher concurs with Hagan's proposition that neither theory holistically accounts for predictability in case outcomes and, instead, finds support for elements of the Structural-Contextual theory.
While Hagan's original framework did not specifically consider terrorism, Smith and Damphousse's application of the theory in a manner that compared terrorists to non-terrorists provided a baseline from which to assess additional variations of datasets that include terrorists (1998). In our results, IE looked more like Smith and Damphousse's terrorist cases, and FR looked more like non-terrorist cases. This is not surprising when one accounts for the types of crimes with which each group is most often charged. Thus, by evaluating two typologies of terror that represent two distinct political environments, we have shown that Structural-Contextual theory operates on a continuum, and that even when two politicised groups are evaluated, sentencing outcomes are relative to one another. The typology with the more proactive environment still has more explained variance, which in this case is due to legal factors.
In previous models, Structural-Contextual theory has focused primarily on explaining variance, but the researcher proposes applying a variation of the theory where proactive political environment explains severity of outcome. In this analysis, we have shown that a proactive political environment not only evidences coupling of the justice system and more explained variance in sentencing, but also predicts more severe conviction rates, sentence lengths and magnitude of resources applied.
One caveat is that within the context of comparing these two typologies of terror, we cannot conclude that a proactive political environment is predictive of the success rate of the primary prosecutorial approach (highest count outcome), since there is no statistical significance in different outcomes between the two. We must also consider that, practically, total conviction rates for FR and IE are both still above 90%, despite a statistically significant difference between the 90.5% rate for FR and the 93.9% rate for IE. We may conclude that for FR, prosecutors can partially compensate for a deficiency in magnitude of preemptive investigatory resources by resourcing efforts at the trial stage. Thus, while there are clearly more preemptive steps taken against IE at the investigatory stage, it would be a mischaracterisation to conclude that the government applies insufficient effort into prosecuting FR cases once they are identified. The differences become more evident in the prosecutorial approach, assessed to have the greatest likelihood of success, as well as at the initial identification/investigatory stages of a case. If strategies were to be changed, there is also no guarantee that a shift to ideological charges would be any more successful against FR than conventional offences, even with the potential for longer sentences.
There were limitations to this study as well. It was not possible to determine how and by what magnitude judges added terrorism enhancements to final sentences, particularly as the mandate to impose these enhancements transitioned from a requirement to a recommendation in 2005. Pre-sentencing recommendations are not publicly accessible. Regional differences and sympathies could also affect both willingness to investigate and willingness to convict/ sentence in different parts of the country (Bradley-Engen et al., 2009, p. 443). Further analysis could evaluate differences in outcomes as they relate to the federal districts in which the cases were filed.
Magnitude of resources could have included other variables if available. Methods such as wiretapping, use of social media exploitation, and FISA requests were not included in the ATS database, as some of them have only relatively recently been applied to terrorism cases, and it is not clear whether the data sources used by ATS would include such information.
Analyses did not account for magnitude of effects beyond highest count variable: i.e. whether the indictee's actions resulted in one death or 100 deaths. While this was partially mitigated by controlling for number of counts, the research could not account for the level of severity of counts beyond the first/most severe due to data availability. Studies such as those by Huff and Kertzer (2017), however, find that severity of violence as a measure of magnitude of effects is ultimately not as significant a factor in the court of public opinion, concluding that type of violence, attributed motivation and the social categorisation of an actor are more significant indicators of a perpetrator being classified as a terrorist than the number of fatalities resulting from any given attack (pp. 62-65). While there is room for magnitude of effects to be explored further, these findings suggest that this variable may be less of an indicator relative to the prosecutorial approach when juries consist of American citizens whose interpretations generally reflect public opinion.
Demographic variables that may illustrate biases or lack thereof, such as religion, were not consistently available in the data for conclusions to be drawn. Similarly, in the two terror typologies in this database, individuals of Middle Eastern or Arab descent would have often been classified as Caucasian, leaving little basis for quantitative comparison against the predominantly white FR indictees. Some studies suggest that parsing this variable would potentially elucidate relevant findings for a research question specific to demographic variables. D'Orazio and Salehyan (2018), for example, find that ingroup-outgroup biases have significantly different outcomes in public opinion surveys, where Arab ethnicity or Islamist group affiliation significantly increase the likelihood of one being labelled a terrorist, as opposed to Caucasian ethnicity or white supremacist affiliation.
Future studies involving the U.S. justice system should investigate the nuances of preemptive approaches to examining domestic terror types, collecting data on additional variables that account for recent technological innovations. This should be evaluated from the perspective of both law enforcement effects and defendant efforts to evade law enforcement advances, as a defendant's manipulation of technology can often be gleaned from court records as well. It would also be valuable to determine a way to code prior criminal record 43 in a manner that would allow it to be used as an additional control for sentencing length. Future studies should incorporate other types of domestic terror, such as Far Left and single-issue terrorists, to determine if there is a continuum of proactive political environment amongst all of the different typologies. In addition, there is value in dissecting the efficacy of the prosecutorial approach between inchoate offences and those crimes that have already been committed, and determining what percentage of cases fall within each category of offence; this too gives evidence of a preemptive approach and diffused prevention strategies (Chesney, 2005;Chesney, 2007). Inclusion of prosecutorial strategies and charges in lower-level courts can add necessary context to how the justice system addresses the threat of terror more holistically. Given the period of time that has passed since 9/11, there is now sufficient data to evaluate pre-and post-9/11 outcomes more comprehensively and in a comparative manner that could be more telling of the scope of the shift in political environment. Finally, it remains to be seen whether the cataclysmic events of FR terror on 6 January 2021, or the rising visibility of FR terror groups in the months and years leading up to it, will serve as an inflection point for the American justice system's treatment of FR criminality. This will need to be examined in much closer detail in future research.
The applicability of themes can be expanded to studies of European justice models, building on Amirault and Bouchard's (2017) study of terrorism sentencing trends in the United Kingdom, or Argomaniz et al.'s (2015) analysis of counter-terrorism approaches in the European Union. The existential questions of liberty and security raised by this field are not constrained to the United States, and this discussion can provide a baseline from which other nation-states dealing with domestic terrorism can seek to evaluate the efficacy of their government's response.

Conclusion
Overall, the results supported the researcher's hypothesis: severity of outcome was of a much higher magnitude for IE. The framework of Structural-Contextual theory was generally supported, which the researcher expected. The results also gave reason to propose a variation of the theoretical framework applicable to between-group comparisons of terrorists.
This study illuminates many compelling trends that contribute to an overarching understanding of how the U.S. government approaches different typologies of terrorism. Political ideology has a significant effect on conviction rates, amplified by the finding that IE offenders are subject to the expenditure of significantly more resources, receive significantly longer sentences, and plead guilty at a significantly higher rate. The data on the threat of FR violence challenges the assumption that IE poses a more existential threat to the security of the United States, yet there is significantly more effort put into ensuring the detection and punishment of IE terrorists. We must consider if this is because IE terrorists are more dangerous or more prolific, or if this represents a reflection of systemic biases within the U.S. political system. This is not to suggest that IE terrorists are not dangerous or that their punishment should be any less severe. Rather, the U.S. political community must collectively reconcile moral consensus in order to consider if it wants to pursue a more equitable approach to identifying and prosecuting domestic terrorism in all its forms, or if it prefers to maintain its historical compartmentalisation of political crimes in an effort to protect other civil liberties.
Notes 1 The Oklahoma City attack represented resurgent right-wing anti-government white nationalism, which had existed within populations across the U.S. since at least the Reconstruction Period (ADL, 2017, p. 2;Lane, 2019;Kaplan, 1995). September 11, 2001 represented the introduction of a foreign threat and Al Qaeda's objective of bringing jihad to the United States. 2 Carter's examination of the foundational elements of right-wing extremism defines it as 'an ideology that encompasses authoritarianism, anti-democracy and exclusionary and/or holistic nationalism ' (2018, p. 174). Far Right terrorism may include religious elements, but this is generally not its sole defining characteristic. 3 Islamic Extremism is a manifestation of religious terrorism. Hoffman describes the core characteristics of religious terrorism as including (1) violence as a sacramental act, (2) the use of religious texts or interpretations to explain contemporary events and justify violence as a legitimate use of force, and (3) an "outsider" oppressor identity that rationalises destructive operations (2006, pp. 88-89). 4 Compared to the Far Left, Chermak and Gruenewald found that Far Right and Al-Qaedaaffiliated groups had similarly high rates of male members and mental health concerns. 5 For a more comprehensive summary of prosecutorial approaches from 1980-1998, see Smith et al. (2002, pp. 312-317). 6 Prior to 9/11, law enforcement was more focused on completed acts of terrorism, as opposed to identifying acts that had not yet been executed (Chesney, 2005, p. 26). The 9/11 Commission Report details how this approach contributed to the failure to anticipate the 9/11 attack (Kean and Hamilton, 2004). 7 This included the use of National Security Letters (NSLs) to obtain information from private companies. 8 The United States Code ("U.S.C.") codifies American federal law. The references herein designate the statutory title, in this case, Title 18 (general criminal violations), and a particular section, in this case, Section ( §) 2332B (dealing with foreign terrorism). 9 For a comprehensive compilation of ISIS sympathisers in the United States, see Vidino and Hughes (2015 ADL (2017, p. 4): '"terrorist groups" as such-i.e., groups that form and exist largely for the purpose of committing terrorist acts-are rare in the United States, where the rule of law is strong and such groups have great difficulties in finding purchase'. 31 Data not available to evaluate Damphousse and Shields' (2007) third category of 'subtle innuendo'. 32 Defendants have the right to a jury trial in criminal cases. Sometimes, the defence will request a bench trial (judge-only, no jury), if the defence believes that a jury trial will result in undue bias against the defendant. In these cases, the judge is the finder of fact. 33 43 Despite the value of prior record, difficulties in recording and coding have resulted in its exclusion from this and many previous studies (Smith and Damphousse, 1996, p. 292;Bradley-Engen et al., 2009, p. 442).     Count Severity=1-29, with 1 being the lowest/least severe and 29 being the highest/most severe   *Magnitude of legal resources was determined by assigning values of 1=yes and 0=no to use of informants, 1=yes and 0=no to use of undercover agents, 1=over and 0=under mean case length in days for case complexity, and 1=over and 0=under mean for number of counts. Each case received a cumulative value of 0-4 based on these four characteristics.