et al. [full author details at the end of the article] Political Behavior https://doi.org/10.1007/s11109-025-10091-x ORIGINAL PAPER Inclusionary and Exclusionary Preferences: A Test of Three Cognitive Mechanisms Marika Landau-Wells1  · Kirsten O. Lydic2  · Joachim Kennedy3 · Benjamin G. Mittman4  · Todd W. Thompson3  · Akhil Gupta3 · Rebecca Saxe3 Accepted: 30 September 2025 © The Author(s) 2025 Abstract Exclusionary social policies take a significant toll on the mental and physical health of targeted groups. Support for specific exclusionary policies does not always align with general antipathy towards the targeted group, however. Does support for specific exclusionary policies rely on particular thought processes (i.e., cognitive mechanisms)? Does opposition? We investigate these questions through the lens of “bathroom laws” across two studies. In Study 1, we use functional neuroimaging to test three candidate cognitive mechanisms from the literature: (1) threat-related emotions (e.g., fear, disgust) supporting exclusionary preferences; (2) mentalizing (e.g., empathy, perspective-taking) supporting inclusionary preferences; and (3) self-regulation (e.g., aligning one’s behavior with one’s goals) supporting inclusion- ary preferences. Consistent with the intergroup conflict and prejudice literatures, we find evidence of a motivated self-regulation mechanism in bathroom law op- ponents. In Study 2, we investigate a possible source of this motivation using text analysis of open-ended policy preference justifications. We find that bathroom law opponents link their policy preference to a small number of specific values, par- ticularly autonomy of action. Taken together, these studies point to a value-driven, motivational account of inclusionary preferences that reconciles puzzling patterns of public opinion, offers new levers for tolerance interventions, and provides some insight into the brain-basis of political behavior. Marika Landau-Wells mlw@berkeley.edu 1 University of California Berkeley, Berkeley, CA, USA 2 University of Pennsylvania, Philadelphia, PA, USA 3 Massachusetts Institute of Technology, Cambridge, MA, USA 4 Case Western Reserve University School of Medicine, Cleveland, OH, USA 1 3 Political Behavior Keywords Political psychology · Neuroscience · FMRI · Gender and politics · Emotions · Self-regulation Introduction Exclusionary social policies take a significant toll on the mental and physical health of targeted groups (e.g., James et al. 2016; Lewis et al. 2015; Patterson et al. 2020; Thapa et al. 2021). Gender identity forms the basis for a variety of exclusionary social policies around the world, from the denial of broad legal protections to the regulation of specific public resources (Hallenbeck et al. 2025). Transgender people are targets of exclusionary policies in a number of countries and the limitations on transgender rights are increasing (TGEU 2025). In the United States, efforts to exclude trans- gender people from access to public resources have intensified over the past decade, with a 300% increase in bills introduced between 2022 and 2024 (Trans Legislation Tracker 2025).1 Proposed and enacted legislation has sought to reduce transgender people’s access to spaces (e.g., public bathrooms), community activities (e.g., youth sports teams), and services (e.g., healthcare), among other restrictions. While nega- tive attitudes towards transgender people and exclusionary policy preferences are correlated (Axt et al. 2021; Flores et al. 2018b), there is significant variation in sup- port for individual policy measures (Lewis et al. 2022, 2024) and tolerant attitudes do not guarantee support for inclusionary policies (Flores et al. 2016; Pew Research Center 2022). The co-existence of support for socially tolerant policies alongside support for exclusionary policies within the same individuals suggests that specific policy preferences are not simply alternative ways to express antipathy or political ideology, but rather result from at least partly distinct cognitive mechanisms.2 In this article, we ask: Does support for exclusionary social policies rely on par- ticular cognitive mechanisms? Does opposition? Answers to these questions have the potential to shape interventions and to illuminate fundamental drivers of political behavior. We examine inclusionary and exclusionary preferences through the lens of “bath- room laws”, which are laws restricting public bathroom access based on an individ- ual’s sex assigned at birth, irrespective of their gender identity. Bathroom access has become the subject of growing debate in the United States (Movement Advancement Project 2025) and elsewhere (e.g., Brooks and Walker 2025; Niazi 2023). Given the nature of this restriction and in line with prior literature (e.g., Kalla and Broockman 2020), we describe bathroom law endorsement as an exclusionary preference and opposition as an inclusionary preference. While a majority of Americans (56–83%) support discrimination protections for transgender people generally, a large minor- 1 We use the term “transgender” as a descriptor for those whose gender identity differs from the sex they were assigned at birth (GLAAD 2022). Following the same guidelines, we use the term “cisgender” to describe those whose gender identity is the same as the sex to which they were assigned at birth. 2 We use the term “cognitive mechanism” consistent with McGraw (2000) to mean “the mental processes underlying judgment and choice” (808). We use the terms “cognitive mechanism”, “thought process”, and “cognitive process” interchangeably to refer to activities in the mind. We use “neural” to refer to activities in the brain. 1 3 Political Behavior ity (41–49%) consistently express support for bathroom laws (Lewis et al. 2023; Luhur et al. 2019; Pew Research Center 2022, 2025). In a 2022 survey, 19% of Amer- icans who supported discrimination protections also supported bathroom laws (Pew Research Center 2022), a pattern that suggests distinct cognitive mechanisms may support bathroom law preferences. Similar patterns have been found in a number of other countries (Flores et al. 2016). We investigate the cognitive mechanisms associated with bathroom law opposi- tion and endorsement with two complementary, preregistered studies.3 In Study 1 (N = 42), we use functional magnetic resonance imaging (fMRI) to measure partici- pants’ neural responses during imagined interactions with transgender people. FMRI is an established method for studying cognitive processes in near-real time (Eickhoff et al. 2018; Kanwisher 2010; Poldrack 2011; Yarkoni et al. 2011) and this approach helps address the concern that people may be unable or unwilling to articulate the thought processes related to their preferences on sensitive issues (Krupnikov et al. 2016). We test three cognitive mechanisms that could plausibly account for bath- room law preferences, each drawn from the existing literature: (M1) the specific, threat-related emotional responses of fear or disgust might underlie exclusionary preferences; (M2) greater mentalizing (e.g., empathy, perspective-taking) might underlie inclusionary preferences; and (M3) greater self-regulation (e.g., aligning one’s behavior with one’s goals) might support inclusionary preferences. Contrary to our expectations, and to the emphasis on disgust in the exclusionary preferences literature (Aarøe et al. 2017; Clifford and Piston 2017; Kam and Estes 2016; Miller et al. 2017; Vanaman and Chapman 2020), we did not find evidence of M1. Also, despite the success of mentalizing exercises as experimental treatments in reducing support for exclusionary policies (Broockman and Kalla 2016; Flores et al. 2021; Kalla and Broockman 2020), we found no evidence of M2. However, we did find a pattern of brain activity consistent with M3, self-regulation. Our results are consistent with the neural correlates of tolerance in the context of racial bias (Amodio and Cikara 2021; Senholzi and Kubota 2016) and with experimental evidence from self-regulation interventions in intergroup conflicts (Halperin et al. 2013; Hurtado- Parrado et al. 2019). Self-regulation requires effort and thus motivation (Inzlicht et al. 2021; Tamir 2016). Our fMRI data provided suggestive evidence that motivation in bathroom law opponents was related to their personal values, but the nature of our data left us no way to test this explanation. In Study 2 (N = 602), we test a value-driven, motivated explanation for inclusion- ary preferences. We re-analyzed data collected by the Nebraska Annual Social Indi- cators Survey (NASIS) (Bureau of Sociological Research 2018),4 which included the same measure of bathroom law preferences used to define the groups in Study 1 as well as an open-ended question asking subjects to provide a justification for their policy preference. We analyzed these open-ended justifications using Schwartz’s theory of basic human values as a framework (Schwartz 1994; Schwartz et al. 2012), 3 Our preregistrations are available on the Open Science Framework. Study 1 (fMRI): https://osf.io/ b8d5h/. Study 2 (text): https://osf.io/m83n7/. D e vi a t i o n s from our planned analyses are noted in the text. 4 The 2018 NASIS data collection period occurred approximately six months before Study 1 and we treat the samples as contemporaneous. 1 3 Political Behavior and a natural language processing (NLP) analytic approach. We hypothesized that valuing autonomy of action and concern for equality would be represented in justi- fications for bathroom law opposition more than alternative value-claims and would be less prevalent in justifications for bathroom law endorsement. Our findings were consistent with both hypotheses, though valuing autonomy of action appears to drive the results. Taken together, our studies suggest that what distinguishes those holding inclu- sionary preferences from those holding exclusionary preferences appears to be a value-driven thought process. Our finding is not simply that values matter, however. Many transgender-related policies – including anti-discrimination measures – appeal to other values (e.g., fairness). Our results show that specific exclusionary policies may map to specific values, which could explain the heterogeneity in patterns of sup- port for transgender-related policies and the partial decoupling from related attitudes (e.g., transphobia). Our investigation into inclusionary and exclusionary preferences makes several contributions. First, our results across both studies support the promise of self-reg- ulation as an intervention strategy. Self-regulation has been used in intergroup con- flict interventions (e.g., Halperin et al. 2013; Hurtado-Parrado et al. 2019) and as a technique for reducing prejudice (e.g., Westerlund et al. 2020). It is largely a learned behavior (McRae et al. 2017) and the motivation to self-regulate can be provided extrinsically through social influence (Tamir 2016). Our fMRI findings also contribute to the study of emotions in politics (for reviews, see Kushner Gadarian and Brader 2023; Webster and Albertson 2022). The direct experience of specific emotions, including fear, anger, and disgust, has been impli- cated in support for far-right political parties (Vasilopoulos et al. 2019), anti-home- less policies (Clifford and Piston 2017), and anti-immigration policies (Aarøe et al. 2017). Real-time measures of emotional response have generally relied on physi- ological indicators (e.g., skin conductance) (e.g., Aarøe et al. 2017; Bakker et al. 2020), but the ability to identify specific emotions with these indicators is low (Yang et al. 2024). We demonstrate how fMRI data can provide this type of discriminability. Finally, we add to the growing body of work on the brain-basis of political behav- ior. There have been several calls for closer integration of neuroscience with political science (Jost et al. 2014; Landau-Wells and Saxe 2020; McDermott 2004; Theodo- ridis and Nelson 2012) and brain-level analyses have already produced insight into the roles of uncertainty and information in candidate evaluation (Haas et al. 2021), the processing of political communications (Casado-Aranda et al. 2022; Moore et al. 2021), the connection between political beliefs and broader neurological processes (Ahn et al. 2014; Schreiber et al. 2022) and brain structures (Nam et al. 2018), as well as informing new models of the influence of partisanship on cognition (Jacoby et al. 2024; Van Bavel and Pereira 2018). But many theories of political behavior rest on as-yet untested assumptions about underlying cognitive mechanisms. Our work demonstrates both the value of testing those assumptions and the importance of complementing neural data with other analytic tools to develop a better understand- ing of political behavior. 1 3 Political Behavior Connecting Preferences, Cognitive Mechanisms, and Brain Data Why do people oppose or endorse exclusionary policies? Research on intergroup atti- tudes has shown that relative warmth or hostility towards a particular outgroup does not always predict preferences for specific policies that may help or harm that out- group (e.g., Hainmueller and Hopkins 2014). The partial de-coupling of intergroup attitudes from related policy preferences may arise from elite cues (e.g., Lenz 2013), fundamental preference instability (e.g., Freeder et al. 2019; Zaller 1992), or the countervailing influence of certain social norms (e.g., Paluck 2009) or worldviews (e.g., Federico and Sidanius 2002). Exclusionary preferences thus require explana- tions beyond simple outgroup antipathy. In the case of the LGBTQ + community, attitudes and related policy preferences have exhibited alignment in some cases and divergence in others. For example, since the early 2000s, an increasing share of the public has expressed acceptance of homosexuality and support for marriage equality (Pew Research Center 2019, 2020). Over the same period, feeling thermometer measures of warmth towards transgender people have also increased, but support for related inclusionary policies generally has remained static (Lewis et al. 2022). Experimentally, researchers have demon- strated that attitudes towards transgender people are more easily altered than trans- gender-related policy preferences (Flores et al. 2018a; Kalla and Broockman 2020, 2023; Lewis et al. 2022), though Flores et al. (2018b) finds that when preferences do change, attitudes change as well. In observational data, Jones and Brewer (2020) find that political awareness affected preferences on bathroom laws as the issue became more salient between 2015 and 2016. Collectively, these studies indicate that bath- room law preferences are influenced by thought processes that do more than channel (non-)prejudicial attitudes. But which ones? And why? Theorized Mechanisms Underlying Inclusionary/Exclusionary Preferences We identified three candidate thought processes (i.e., cognitive mechanisms) in the literature, each of which could account for either inclusionary or exclusionary preferences. The first candidate mechanism comes from the threat perception literature. Threat perception has been linked to exclusionary policy preferences related to immigration (Aarøe et al. 2017; Hainmueller and Hopkins 2014; McLaren 2003), LGBTQ + com- munities (van Leeuwen et al. 2023; Vanaman and Chapman 2020), and race relations (Bobo 1983; Kinder and Sears 1981). Bathroom laws are theorized to address two perceived forms of threat: physical safety threats, particularly for cisgender women and children (Barnett et al. 2018; Michelson and Harrison 2020), and literal or met- aphorical purity threats (Miller et al. 2017; Vanaman and Chapman 2020). These perceptions of threat are associated with the specific emotional responses of fear (e.g., Michelson and Harrison 2020) and disgust (e.g., Vanaman and Chapman 2020), respectively. Notably, threat perceptions (and emotional associations) attach to group members in any context, not simply in bathrooms (e.g., Cottrell and Neuberg 2005). Nor does threat perception need to reflect real risk to influence preferences (e.g., Hasenbush et al. 2019; Hopkins et al. 2019; Kinder and Sears 1981; Landau-Wells 1 3 Political Behavior 2018). Nevertheless, according to this theorized mechanism, the emotional responses of either fear or disgust during interactions with transgender people would be associ- ated with exclusionary preferences (M1). Our second candidate mechanism comes from the literature on intergroup contact interventions (for a review, see Paluck et al., 2019). Specifically, interventions on transgender attitudes and related policy preferences have found that exposure to the personal narratives of transgender people in certain settings can durably alter policy preferences, including reducing support for bathroom laws (Broockman and Kalla 2016; Flores et al. 2021; Kalla and Broockman 2020). The effectiveness of these treatments is theorized to be driven by mentalizing – that is, by encouraging subjects to consider thoughts, feelings, and beliefs that are different from their own (Kalla and Broockman 2023).5 According to this theorized mechanism, mentalizing during interactions with transgender people would be associated with inclusionary prefer- ences (M2). Our third candidate mechanism comes from the literatures on prejudice and inter- group conflict. Multiple studies have identified self-regulation – a broad set of inter- nal cognitive processes that align one’s behavior with one’s goals (Inzlicht et al. 2021; Tamir 2016) – as improving social tolerance of outgroups (for reviews, see Amodio and Cikara 2021; Ochsner and Gross 2013; Senholzi and Kubota 2016). While self-regulation can occur in many ways, cognitive reappraisal, which involves actively adjusting one’s response to a stimulus, has received significant focus (for a review, see Gross 2015). Differences in spontaneous self-regulation arise from dif- ferences in motivation (Tamir 2016). Motivation can come from intrinsic pressures (e.g., adhering to one’s personal values) or extrinsic pressures (e.g., adhering to social norms) (e.g., Mattan et al. 2018). Experimentally induced (extrinsic) reappraisal led to small but significant increases in support among Israelis for conciliatory poli- cies towards Palestinians (Halperin et al. 2013), support for conciliatory statements about rebel groups in Colombia (Hurtado-Parrado et al. 2019), and tolerance towards migrants in Finland (Westerlund et al. 2020). This theorized mechanism suggests that self-regulation during interactions with transgender people would be associated with inclusionary preferences (M3). Measuring Cognitive Mechanisms While the literature thus identifies several possible thought processes that might explain specific exclusionary policy preferences, their measurement presents chal- lenges. M1 (threat-relevant emotions) requires accurately capturing real-time emo- tional responses, which standard self-report measures cannot do reliably (Kushner Gadarian and Brader 2023). M2 (mentalizing) requires measuring an activity that often occurs without conscious reflection and a prompt to self-report the activity may simply induce it. M3 (self-regulation) requires measuring an intermediate, fast, iterative process, but retrospective self-reporting generally captures only end-states (i.e., the final feeling one has). A concern with respect to all three mechanisms is that 5 We use the term “mentalizing” consistent with the cognitive science literature to capture a broad range of inferences about other minds (Fehlbaum et al. 2022; Frith and Frith 2006; Saxe and Kanwisher 2003). 1 3 Political Behavior people may be unable or unwilling to articulate their thought processes on sensitive issues. Functional magnetic resonance imaging (fMRI) offers one way to address these challenges and is an established method for investigating cogntive processes (Eick- hoff et al. 2018; Kanwisher 2010; Poldrack 2011; Yarkoni et al. 2011). An fMRI study collects a sequence of three-dimensional images that capture the flow of oxy- genated blood within the brain as a proxy for neural activity while a subject performs a task (Poldrack et al. 2011).6 Increased oxygenated blood reveals locations in the brain that are relatively more active at specific times within the task. FMRI collects data in near real-time and so can capture intermediate processes during tasks, as well as capturing responses outside the subject’s control. Without delving too far into the technical details, the remainder of this section introduces several aspects of fMRI study design and data that are relevant for our investigation of cognitive mechanisms. To begin with, most fMRI studies rely on a within-subject design. A single partici- pant in an fMRI study often will be exposed to multiple conditions many times (for example, Condition A = reading scary stories and Condition B = reading emotionally neutral stories). A subject’s brain responses during each instance of each condition are often summarized into a single condition-level average image. These summary images can then be compared in several ways to identify the pattern of brain activity associated with each condition. Because the brain is spatially organized, these pat- terns can be mapped onto a standardized brain template to identify the brain areas that are relatively more active during each condition (Eickhoff et al. 2018). This method of associating brain areas with cognitive process on the basis of performing a directly relevant task is known as forward inference (Hutzler 2014; Poldrack and Yarkoni 2016). Complex cognitive processes rely on distributed combinations of neural activ- ity from many brain areas, however, so the mapping of brain areas to complex cogni- tive functions is not one-to-one (Fedorenko et al. 2013; Shine and Poldrack 2018).7 This means that reverse inference – asserting that a particular cognitive process is at work given the observed activity of a specific brain area – must be done with great care (Hutzler 2014). These features of fMRI study data affect our approach to measuring cognitive mechanisms in several ways. First, we measure mechanisms as patterns of brain activ- ity captured in three-dimensional images and these images constitute the data for our analyses in Study 1. Each image is made up of three-dimensional pixels, known as voxels, which are 2–3 mm 3. Voxels have fixed locations in a standard brain template, expressed as (x, y, z) coordinates, and can represent the activity of as many as one million neurons (Huettel et al. 2004). Like a grayscale image, voxels not only have a location, but also a brightness value that can vary between images. These values are often the result of a statistical models (e.g., beta values) or statistical tests (e.g., t-statistics) (Poldrack et al. 2011). Because there are thousands of voxels in a brain image, thresholds of statistical significance take into account both the large number 6 This is known as the blood-oxygen-level-dependent (BOLD) signal. In a standard fMRI scan, one image of the brain is collected approximately every two seconds. For a longer introduction to neuroscience methods for political scientists, see Haas (2016) and Landau-Wells (2024). 7 There is no single “fear center” of the brain, for example (Zhou et al. 2021). 1 3 Political Behavior of tests and the spatial dependence of many voxels. Voxels are spatially dependent because brain areas are much larger than a single voxel. As a result, brain activity is often analyzed based on clusters of active voxels whose spatial locations map on to one or more brain areas. These clusters are also known as regions-of-interest (ROIs).8 Second, because a single condition-level summary image for a single subject is not generally interpretable on its own, we must use comparisons between conditions to isolate the brain activity of interest. We use two types of between-condition compari- son in our study of cognitive mechanisms. The first is a large-scale contrast, which compares two conditions by subtracting the average activity in one from the other in a voxel, cluster, or ROI (Poldrack et al. 2011). Contrasts are expressed as: Condition A (e.g., reading scary stories) > Condition B (e.g., reading neutral stories) and are the basis of the “brain blob” images so often associated with fMRI. The second type of between-condition comparison we use is to examine small-scale, voxel-level pattern similarity. This is known as representational similarity analysis (RSA) (Haxby et al. 2001; Kriegeskorte et al. 2008). RSA is a more fine-grained method of assessing dif- ferences between conditions using measures of similarity applied to vectors of voxel values (e.g., correlations). In both cases, the analytic goal is to generate interpretable, subject-level patterns of neural response, which can then be tested across the study (or within a subgroup) for consistency. The third consequence of the nature of fMRI data for our measurement of cogni- tive mechanisms is the challenge of making inferences about cognitive processes that we do not directly induce (i.e., reverse inference). Our goal is to detect the spon- taneous use of our three candidate cognitive mechanisms during imagined interac- tions with transgender people. This requires a strong prior expectation of what each mechanism would look like if it were present in our data. We deal with this inferential problem in two steps. First, we use prior neuroimaging studies to define ROIs where we have a non-zero probability of observing each mechanism if it is at work. All of our candidate mechanisms have been studied extensively with fMRI and we lever- age over 80 previously published neuroimaging studies with approximately 2,250 participants to create four “search spaces” for our own study (Fear, Disgust, Men- talizing, and Self-Regulation), which we detail in the next section. Our second step is to specify how we test for each cognitive process within its search space. As our preregistration details, these tests vary based on the candidate mechanism. Broadly speaking, we use RSA in our test of M1 due to the relatively high degree of resolution needed to distinguish between fear and disgust and their overlapping search spaces. We use ROI-level contrast results for our tests of M2 and M3. We provide the specif- ics of all testing strategies and thresholds for inference in the Analytic Design section of Study 1. Generating Search Spaces for Candidate Mechanisms We used two approaches to generate our search spaces from the prior literature. Our approach for three of the four search spaces was to use coordinate-based meta-analy- sis (CBMA), a process for deriving a map of concordant findings across many fMRI 8 Some regions-of-interest correspond to cleanly defined anatomical structures, but often they do not. 1 3 Political Behavior studies of the same cognitive process (Eickhoff et al. 2009; Yarkoni et al. 2011). CBMA concordance maps (hereafter, CBMA maps) are three-dimensional images indicating clusters of voxels where the there is a non-zero probability of brain activ- ity during the cognitive process captured by the input studies. It is important to note that CBMA maps do not reflect “networks”; rarely does an input study report a pat- tern of activity that maps perfectly onto the concordance map. Rather, CBMA maps offer a principled way to apporach the problem of reverse inference by pre-defining the space where a cognitive process might be observed (Hutzler 2014).9 Appendix C provides a full description of our search space definition procedures. In the case of M1 (threat-relevant emotions) we conducted one CBMA to generate a search space for the experience of disgust and another to generate a search space for the experience of fear. Our Disgust CBMA included 24 studies comprising 687 subjects and yielded a concordance map of six clusters of voxels with a non-zero probability of activation for the experience of disgust (Table C4). Our Fear CBMA included 26 studies comprising 740 subjects and yielded a concordance map of seven clusters of voxels with a non-zero probability of activation for the experience of fear (Table C5). In the case of M3, we conducted a single CBMA to generate a search space for self-regulation. Our Self-Regulation CBMA included 28 studies compris- ing 1,047 subjects and yielded a map of ten clusters of voxels with a non-zero prob- ability of activation during self-regulation (Table C3). We used a different approach to generate a search space for mentalizing (M2). Mentalizing is associated with the reliable co-activation of a network of seven ROIs (Dufour et al. 2013). Activation across this network can be induced by a short scanner task known as a functional localizer (Dodell-Feder et al. 2011). In Study 1, we had participants complete this localizer task after the main experiment. In all cases, our search space definition is independent of the data used for our main analysis, which allowed us to avoid circular reasoning when making inferences about the role of specific brain areas in cognitive processes (Poldrack and Mumford 2009). Hypotheses for Each Cognitive Mechanism With these search spaces in mind, we formalized the expectations for each cognitive mechanism in a series of hypotheses. To distinguish between the expectations associ- ated with exclusionary preferences and inclusionary preferences, we express each hypothesis with respect to the group to which it applies. “Endorsers” are those who support bathroom laws; “Opponents” are those who do not support such laws. First, we consider whether threat responses, operationalized as the specific nega- tive emotions of either fear or disgust, are associated with exclusionary preferences (M1). Based on the balance of the existing political psychology literature in favor of a disgust response, we hypothesized that in the Disgust and Fear Search Spaces: 9 For an in-depth discussion of the use of CBMAs for political science, see Landau-Wells (2024). 1 3 Political Behavior H1a Endorsers’ neural responses during interactions with transgender people in neu- tral scenarios will more closely resemble responses to interactions with cisgender people in disgust-inducing scenarios than in either fear-inducing or neutral scenarios. H1b Opponents’ neural responses during interactions with transgender people in neutral scenarios will more closely resemble responses to interactions with cisgender people in neutral scenarios than in either fear- or disgust-inducing scenarios. H1c Opponents and Endorsers will not exhibit the same degree of similarity between their neural responses to the transgender-neutral and cisgender-neutral conditions. Second, we consider whether mentalizing during imagined interactions with trans- gender people is associated with inclusionary preferences (M2). Based on the inter- group contact literature, we hypothesized that in the Mentalizing Search Space: H2a Opponents’ neural responses during interactions with transgender people in neutral scenarios will show greater activation than during interactions with cisgender people in neutral scenarios. H2b Opponents and Endorsers will differ in their neural responses during imagined interactions with transgender people in neutral scenarios (relative to interactions with cisgender people in neutral scenarios). Third, we consider whether self-regulation is associated with inclusionary prefer- ences (M3). Based on the prejudice and intergroup conflict literature, we hypoth- esized that in the Self-Regulation Search Space: H3a Opponents’ neural responses during interactions with transgender people in neutral scenarios will show greater activation than during interactions with cisgender people in neutral scenarios. H3b Opponents and Endorsers will differ in their neural responses during imagined interactions with transgender people in neutral scenarios (relative to interactions with cisgender people in neutral scenarios). Of these, all were preregistered except H2a due to an oversight. Additional preregis- tered hypotheses, tests, and results are presented in Appendix E. 1 3 Political Behavior Study 1: Neural Correlates of Policy Preferences Our fMRI study took place in Massachusetts three months after a bathroom law had appeared on the ballot.10 Participants were recruited from the Boston area using online advertisements and posted flyers. Our recruitment materials described an fMRI study on social interaction with no mention of transgender people or politics. Interested individuals completed an online questionnaire to assess their eligibil- ity. In addition to standard MRI safety screening questions, they answered ten ques- tions about their preferences on political issues. Each response was recorded along a 5-point Likert scale (“Strongly Disagree”, “Disagree”, “Neither Agree nor Disagree”, “Agree”, “Strongly Agree”). Nine of the policies were unrelated to the study and served to mask our interest in bathroom laws. All participants who indicated that they agreed or strongly agreed with the statement “Transgender people should be required to use the public restrooms of the gender they were born into” were coded as Endorsers. Those who disagreed or strongly disagreed with the statement were coded as Opponents.11 Those who “neither agreed nor disagreed” were not eligible for the study. Eligible participants came to the lab for a single fMRI scanning session. Each par- ticipant was paid $50 for the 2-hour visit. Prior to the scan, participants were informed that the experiment involved the investigation of emotional responses to social inter- actions and that they might experience negative emotions. At that point, participants completed a standard consent form for fMRI research. All consenting participants were scanned. In the post-scan debrief, we informed participants about the study’s hypotheses and gave them the option to contribute their data to the study or not. Full compensation was assured in either case. Participants were left in private to review the data consent form. All participants who received the debrief also consented to contribute their data to the study. All experimental procedures were approved by our institutional review board. The double-consent procedure was unusual but, as we explained to participants, we could not disclose the exact nature of the study without compromising the data generating process itself. We provide technical details of our study design and analysis procedures in Appendices A, B, and C. Sample Our sample consisted of 43 right-handed adults (mean age = 26.0; 20 women) with no history of brain damage or diagnosed mental health disorders and normal or cor- rected-to-normal vision. All participants were U.S. citizens for whom English was a native language. Participants were pair-matched across groups on gender, age, and education level. 21 participants indicated support for bathroom laws in the screening survey (Endorsers). 22 participants opposed bathroom laws (Opponents). Based on relevant neuroscience literature, we preregistered a minimum sample size of 20 per group with a stopping criterion based on preprocessed data quality that allowed for 10 Question 3 in the November 2018 election. 11 Our wording followed the format used by the Pew Research Center, e.g., Pew Research Center (2022). 1 3 Political Behavior the possibility we would collect a small number of additional datasets after reaching our minimum threshold. One participant with otherwise viable data was excluded from all analyses because their match failed our data quality screen and our stopping criterion had been reached. Our total matched sample thus included 42 subjects. Design Prior to the start of their scanning session, participants completed a short paper ques- tionnaire to gather their views on three social groups (transgender people, immi- grants and refugees, and evangelical Christians). The questionnaire served to remind participants of transgender people as a social category without explicitly stating the purpose of the study. Figure 1 illustrates the Imagined Social Interaction Task (ISIT), which partici- pants completed in the scanner over the course of eight short scanning runs (four pri- mary Reading runs; four attention check Testing runs). A “run” refers to a continuous period of data collection during which participants perform tasks and, crucially, must lie still. Runs ranged from 3.5 to 6.5 min long, with the option for rest in between. During the four primary Reading runs, participants read short stories describing interactions with unfamiliar people in three types of scenarios: (1) fear-inducing sce- narios; (2) disgust-inducing scenarios; and (3) emotionally neutral scenarios. Each scenario was written in the second person (“You are…”). The unfamiliar person was represented by a photo presented above each story. Photos were either cisgender men or transgender women. Our use of images to convey gender identity is consistent with research demonstrating that pictures generate stronger responses in the study of transgender attitudes than text alone (Axt et al. 2021). All story and face stimuli were standardized on a number of dimensions and their effectiveness was validated in separate samples (see Appendix B). We used only those stories that effectively Fig. 1 The Imagined Social Interaction Task Condition Examples and Run Structure. Note: Faces blurred for publication purposes 1 3 Political Behavior evoked one of the three target emotional states (calm, fear, disgust). Table 1 provides examples of each story condition. Participants were instructed to read these stories and to vividly imagine themselves interacting with the people whose faces were pro- vided. An attention check Testing run followed each primary Reading run. Our condition of interest was the imagined interaction with transgender women in neutral scenarios (T). Our three comparison conditions were interactions with cisgender men in: fear-inducing scenarios (F), disgust-inducing scenarios (D), and neutral scenarios (C). Each story/face presentation lasted for 16 s and constituted one trial. Each of the four primary runs consisted of sixteen total trials (four conditions presented four times, quasi-randomized), as shown in Fig. 1. Based on this design, we were able to identify participant-level brain responses to imagined interactions in our four experimental conditions (T, D, F, C) with 12–16 realizations per condi- tion, depending on data quality. We assessed data quality as a function of both head motion, which renders fMRI data too noisy to be useful, and participant attention (see Appendix A). Overall, 98% of the data collected from our 42 participants during the ISIT was retained for the main analyses. By focusing on the response to intergroup contact, rather than to group-relevant messages, our paradigm differs from previous research into the brain-basis of politi- cal behavior focused on belief- or identity- congruent/incongruent information pro- cessing (for a review, see Bakker and Schumacher 2025). While intergroup contact may generate a similar valence-based difference in response to stimuli (i.e., pleasant/ unpleasant), prior work suggests we cannot assume such a dichotomy in advance (e.g., Bruneau et al. 2012). Analytic Approach For each mechanism, we confined our analyses to the corresponding search spaces. In the case of M1 (threat-related emotions), prior fMRI research indicated that we would need to account for the fact that some brain areas are active in the process- Table 1 Short story stimulus examples Story Example Calm Fear Disgust Type Ratings Ratings Ratings Neutral You have decided to see a movie in the middle of 84.2 (17.8) 4.9 (14.9) 4.6 (13.8) the day. The person at the ticket counter asks what you would like to see. You spend a few moments considering your options. Then you buy a ticket and go into the theater. Fear You fall asleep in your bed and wake up some 5.0 (15.1) 91.4 (18.5) 28.5 (17.2) place dark and damp. You realize you are lying in a grave. A person is standing above you, holding a shovel. The person begins to bury your face and body with loose dirt. Disgust You are on a plane. The person next to you looks 15.0 (20.5) 25.1 (29.7) 92.4 (16.4) queasy and reaches for the airsickness bag. Before they can use it, the person vomits all over the seat in front. Chunks of vomit splash onto you and the smell is very strong. Note: Ratings are means (0–100 scale) with standard deviations in parentheses 1 3 Political Behavior ing of both disgust and fear (Kragel and LaBar 2016; Nummenmaa and Saarimäki 2019). Consistent with this, our Fear and Disgust search spaces overlapped in several brain areas (e.g., left and right amygdala). To refine the ROIs for our tests of M1, we used data from our Fear (F) and Disgust (D) conditions to run voxel-level similarity analyses within each of the ROIs identified by each CBMA. We looked for ROIs in which the voxel-level pattern of activity for a given emotion (fear or disgust) could be distinguished from: (1) the cisgender-neutral condition (C), and (2) the other nega- tive emotion. This procedure ensured we would have the ability to detect each emo- tional response during our condition of interest (T) in our main analyses. We deviated slightly from our preregistration in the similarity measures used at this stage (and, therefore, in testing H1a-H1c) due to unforeseen incompatibilities between our data preprocessing and analysis pipelines (see Appendix A). Our revised similarity mea- sure in all cases is Pearson’s correlation. Our emotion detection results are presented in Appendix C. In brief, we found three ROIs in which we could reliably distinguish fear from the other two conditions (left medial occipital gyrus, right amygdala, right inferior temporal gyrus) (Table C6) and six ROIs in which we could reliably distin- guish disgust (Table C7).12 This reduced set of ROIs thus formed the search space in which we tested M1. Our hypothesis tests compare similarity scores between our four conditions in these ROIs using robust t-tests for dependent samples. We treat the similarity test in each ROI as an individual test following Rubin (2021) with an unadjusted alpha value. We use unadjusted alphas across CBMA-identified ROIs for two reasons. First, CBMA maps do not come with an expectation of co-activation across all ROIs in the map or any particular combination. Instead, the particular configuration of clusters is expected to be somewhat task- and study-dependent. For example, the median number of active clusters in the studies contributing to our Disgust CBMA is three (out of 6 in the map) and no study activated all clusters. Second, due to the differing number of clusters across CBMA maps, we found it difficult to justify using different standards of strin- gency for the same stimuli. With these considerations in mind, we chose to test and interpret results on a region-by-region basis, though we also report the results with adjusted alphas where the less conservative threshold is statistically significant. In the case of M2 (mentalizing), we were able to identify all seven ROIs within the mentalizing network for 41 out of 42 participants (Table C1) using the local- izer task (Dodell-Feder et al. 2011). There were no differences in activation between our experimental groups during the localizer task (Table C2). Because the mental- izing network is characterized by co-activations of its ROIs, we treat mentalizing as an all-or-nothing proposition where evidence in support of the mechanism must be network-wide activation. This amounts to a conjunction test across ROIs, which also does not require alpha adjustment (Rubin 2021). Nevertheless, we report results with an adjusted alpha for reference where ROI-level findings are statistically significant. Our analytic approach to testing M3 (self-regulation) focused on ROI-level con- trasts using robust one-sample, one-sided t-tests. As with M1, we interpret each test as ROI-specific, since we had no a priori expectation or clear threshold for how to 12 Our Disgust search space included portions of: left and right amygdala, left and right insula/inferior frontal gyrus, left parietal lobule, and right claustrum. 1 3 Political Behavior interpret the findings across the CBMA map. In our CBMA studies, the median num- ber of active clusters was two and no study activated all ten clusters. As with M1, we report results with adjusted alphas for context. We also report the extent to which the patterns of activation we observe in our data resemble those in the studies included in the Self-Regulation CBMA. Results M1: Fear or Disgust Experience Contrary to H1a, we did not find evidence that the group holding exclusionary preferences (Endorsers) experienced clearly defined negative emotions during their imagined interactions with transgender women. Within our Disgust search space (Table D1) and our Fear search space (Table D2), we did not find any ROI in which the transgender-neutral condition significantly resembled either a disgust or a fear response.13 We provide detailed discussion of our testing methods and present all test results in Appendix D. We also found no evidence of H1b. In the group holding inclusionary preferences (Opponents), there was no ROI in either the Disgust search space (Table D3) or the Fear search space (Table D4) in which the representation of the transgender-neutral condition was significantly more similar to the cisgender-neutral condition than to the negative emotion conditions. Finally, we found no clear evidence for H1c. Across all CBMA-identified ROIs, we found no regions with any significant group difference in the representation of the transgender-neutral condition relative to the representation of the cisgender-neutral condition (Tables D5 and D6). In sum, we found no evidence that similarity to either a fear or a disgust response during imagined interactions with transgender women was associated with bathroom law preferences. M2: Mentalizing Contrary to H2a, we did not find evidence of greater mentalizing in those holding inclusionary preferences (Opponents) during their imagined interactions in the trans- gender-neutral (versus cisgender-neutral) condition (Table D7).14 Fig. 2a displays the patterns of relative activation in each of the seven Mentalizing ROIs for each group. The search space itself is depicted in Fig. 2b. The y-axis of Fig. 2a displays the T > C contrast value. Only two ROIs – dorsal medial prefrontal cortex (dmPFC) and mid- dle medial prefrontal cortex (mmPFC) – show greater contrast values in Opponents (Table D7).15 Both ROIs also maintain statistical significance using a Bonferroni- 13 Manipulation checks reported in Appendix E demonstrate that our stimuli successfully induced both fear and disgust as intended. 14 H2a was not preregistered. We used the same approach for testing H2a as we did for H1a (preregis- tered), matching the hypothesis testing method to the method used to define the search space itself. 15 For Opponents: dmPFC t(20) = 5.14, p < .001; mmPFC t(20) = 2.96, p < .01 1 3 Political Behavior Fig. 2 Mentalizing search space patterns of activation (left) and map (right). Note: (a) Distribution of the T > C contrast by group with significance testing results for one-sided, one-sample t-tests: * p < 0.05; ** p < 0.01; *** p < 0.001. (b) Visualization of the seven ROIs within the mentalizing network derived from ROIs identified in Dufour et al. (2013) overlaid on a glass brain using MRIcroGL (1.2.2) adjusted alpha threshold.16 Endorsers showed no significant contrast differences in any ROI (Table D8).17 Our methods and full test results are included in Appendix D. A test of group differences (H2b) indicated a significant difference (not only a difference in significance) in dmPFC ( t (38.2) = 3.01, p < 0.01), but not mmPFC ( t (36.3) = 1.81, p = 0.08). See Table D9 for full results. We validated that the dif- ference in dmPFC was related to the transgender-neutral interaction specifically by examining the disgust and fear conditions in dmPFC, also contrasted against the cis- gender-neutral condition. We found no group difference during either emotion condi- tion, which suggests that the difference we observed was related to some aspect of the transgender-neutral condition specifically.18 In sum, we found no evidence of the type of network-wide activation that would suggest greater mentalizing activity distinguishes those with inclusionary preferenc- es.19 We did find group differences in dmPFC responses, however. DMPFC acti- vation is associated with a wide variety of mental states and cognitive processes, including estimates of self-relevance, particularly the relevance of a stimulus to one’s self-perception and identity (Berkman et al. 2017; Lieberman et al. 2019), as well 16 The mentalizing network consists of seven ROIs, suggesting a simple Bonferroni-adjusted alpha of 0.05/7 = 0.007. The test in dmPFC yielded a p-value of less than 0.001 and the test in mmPFC yielded a p-value of 0.004. 17 For Endorsers: dmPFC t(19) = 0.54, p = 0.59; mmPFC t(19) = 0.01, p = 0.99. 18 Group difference for D > C in dmPFC (two-sided robust, bootstrapped p-value): t = 0.65, p = 0.52. Group difference for F>C in dmPFC (two-sided parametric): t(34.3) = −1.06, p = 0.29. 19 All conclusions remain the same if we drop the unmatched participant and use a matched sample of 40. 1 3 Political Behavior as self-reflection upon certain personal values (Teed et al. 2020). Due to Study 1’s design, however, it is not possible for us to demonstrate that any of these processes is at work. Therefore, we consider the implications of dmPFC activity within Study 2. M3: Self-Regulation As with M2, we tested M3 using the mean T > C contrast in each ROI within the Self- Regulation search space. Consistent with H3a, those holding inclusionary prefer- ences (Opponents) showed significantly greater activity during imagined interactions with transgender women (versus cisgender men) in neutral scenarios in seven out of ten ROIs in our Self-Regulation search space (Table D10).20 Endorsers do not show a similar pattern (Table D11). Figure 3a illustrates the patterns of activation within the Self-Regulation search space for both groups. Figure 3b visualizes the search space. While Opponents show greater activation across the majority of regions in the search space, we found only one ROI in support of H3b, where there is a significant difference (not only a dif- ference in significance) between groups (left inferior frontal gyrus, t (23.9) = 1.81, p < 0.05). Results for all ROIs are provided in Table D12. Our methods are dis- cussed in detail in Appendix D. As there is no gold standard for judging the similarity of a single study to a CBMA map, we examined studies within the self-regulation CBMA exhibiting similar pat- terns of activation. Our results in Opponents bear the greatest similarity to Silvers et al. (2015), differing only in the reported activation of one additional ROI in our study Fig. 3 Self-regulation search space patterns of activation (left) and map (right). Note: (a) Distribution of the T > C contrast by group with significance testing results for one-sided, one-sample t-tests: * p < 0.05; ** p < 0.01; *** p < 0.001. (b) Visualization of the ten ROIs within the self-regulation CBMA map overlaid on a glass brain using MRIcroGL (1.2.2) 20 The Self-Regulation CBMA consists of ten ROIs, suggesting a simple Bonferroni-adjusted alpha of 0.05/10 = 0.005. Only the test in left IFG yielded a p-value of less than 0.005. 1 3 Political Behavior (left superior frontal gyrus). Notably, Silvers et al. (2015) investigate both instructed and spontaneous emotion regulation, which they conclude are “highly overlapping” (15). We consider this to be directional support for our inference that self-regulation occurs in Opponents in Study 1. Thus, across our brain-level tests of three cognitive mechanisms that could sup- port exclusionary or inclusionary preferences, we found the strongest evidence for a self-regulation mechanism (M3), with additional evidence for the involvement of dorsal medial prefrontal cortex (dmPFC). Study 1 cannot, however, address the question of why we observed these differ- ences between those holding inclusionary versus exclusionary preferences. In the self-regulation literature, motivation plays a crucial role in explaining when regula- tion occurs because regulation requires effort (Inzlicht et al. 2021; Silvers and Guassi Moreira 2019; Tamir 2016). Motivation to self-regulate can result from intrinsic pres- sures (e.g., adhering to one’s personal values) or extrinsic pressures (e.g., adhering to social norms) (e.g., Mattan et al. 2018). DMPFC activity suggests one possible source of intrinsic motivation for bathroom law Opponents: attending to their self- perception, identity, and values (Berkman et al. 2017; Lieberman et al. 2019; Teed et al. 2020). These associations are speculative, however, and cannot be directly addressed without knowing more about the justifications people have for the policy preferences they hold. Study 2: Preference Justifications and Personal Values To probe the question of motivations for inclusionary preferences raised by Study 1, we turned to the 2018 Nebraska Annual Social Indicators Survey (NASIS). The 2018 NASIS data were collected at approximately the same time as Study 1 and used the same wording to measure bathroom law preferences (Bureau of Sociological Research 2018). Additionally, the 2018 NASIS collected open-ended responses to a question asking respondents to justify their preferences. After applying our preregis- tered exclusion criteria, we identified 602 valid text justifications of which 311 were authored by Opponents and 291 by Endorsers.21 Based on Study 1, we posited that specific personal values held by Opponents could account for both their stance against bathroom laws and greater self-regula- tion during our fMRI study by acting as an intrinsic source of motivation. We used Schwartz’s theory of basic human values as a framework for identifying personal val- ues that might provide motivation in the domain of bathroom laws (Schwartz 1994; Schwartz et al. 2012). Schwartz’s theory specifies four higher-order values (Self- Transcendence, Openness to Change, Self-Enhancement, and Conservation) under which a number of “value categories” are clustered. We preregistered two value-cat- egories – Self-Direction (Action) and Universalism (Concern) – as those most likely to capture the motivational logic we wished to test. The value-category of Self-Direc- tion (Action) includes valuing the ability to choose one’s own goals, independence, 21 The data were shared with our group upon request. We used responses from Question 37 to identify Opponents and Endorsers and the corresponding text from Question 38 in our analyses. 1 3 Political Behavior freedom of action, and privacy. Freedom of action and privacy seemed particularly relevant in the case of bathroom law preferences. The value-category of Universal- ism (concern) includes valuing equality, being just, and having a world at peace. Values of equality and justness seemed relevant to the issue of equal access to a basic necessity. Thus, with Study 2 we ask: do people justify their opposition to bathroom laws using language that seems to express one or both of these two value-categories? All twenty value-categories and sub-category values are provided in Table F2. Previous research into the relationship between values and bathroom law prefer- ences has relied on researcher-driven categorization and coding of text (Cox et al. 2022; Kazyak et al. 2021). This process can introduce confirmation bias, however (Chakrabarti and Frye 2017), and is often not replicable (e.g., Mikhaylov et al. 2012). To mitigate these concerns, we used natural language processing tools to categorize the values expressed in the 2018 NASIS dataset. Specifically, we estimated the val- ues expressed in the 2018 NASIS bathroom law justifications by measuring their semantic similarity to a benchmark dataset of justifications for other policy propos- als (i.e., ending affirmative action) already hand-annotated for an extended version of Schwartz’s framework. These benchmark annotated justifications are a subset (N = 2,655) of the Human Values Behind Arguments (HVBA) Dataset (Kiesel et al. 2022). Within this subset, 181 justifications expressed our two target value-categories of interest and 2,474 expressed the eighteen other value-categories. For simplicity, we refer to the two target value-categories as “target values” and the remaining value- categories as “other values”. Additional information is provided in Appendix F. Table F1 compares the demo- graphic characteristics of our fMRI study sample and the 2018 NASIS sample. Analytic Approach To compare our two datasets of text-based policy justifications, we used a frozen, pretrained, open-source large language model (LLM) as a common semantic space. Our model of choice was a transformer-based sentence-encoder fine-tuned on the task of semantic similarity (stsb-mpnet-base-v2) (Reimers and Gurevych 2019).22 This particular model encodes arbitrary length text into a 768-dimension dense vector space and out-performs both traditional word-level embedding models and many commercially developed encoders (e.g., OpenAI’s text embedding models) on semantic similarity benchmarking tasks (Muennighoff et al. 2023). Because the model is frozen and pretrained, its embedding space has not been influenced by our preferred outcomes and the embeddings themselves are replicable. The model is also small enough to run locally and thus does not require purchasing additional compu- tational resources. We encoded each dataset for a total of 602 NASIS justifications for bathroom law preferences and 2,655 annotated justifications covering a variety of other policy issues. We measure the semantic similarity between each bathroom 22 The model is available in the SentenceTransformers Python module (Reimers 2022) and via the Hug- ging Face platform: h t t ps : / / hu g gi n g f ac e . c o/ s en t e n ce - t r an s fo r m e rs / s t sb - mp n e t -b a s e -v 2. 1 3 Political Behavior law stance justification and each values-annotated benchmark justification using the cosine similarity of their two embedding vectors.23 Using the same definition of Opponent and Endorser as Study 1, we preregistered the following hypotheses prior to data analysis: H4a On average, the justifications offered by bathroom law Opponents in the 2018 NASIS dataset will be more semantically similar to justifications in the HVBA bench- mark dataset annotated for Self-Direction (Action) and Universalism (Concern) than those annotated for the other eighteen value-categories. H4b On average, Opponents’ justifications in the 2018 NASIS dataset will be more semantically similar to justifications annotated for Self-Direction (Action) and Uni- versalism (Concern) in the HVBA dataset than will Endorsers’. In both cases, we test our directional hypotheses with one-sided t-tests (paired in the first case, unpaired in the second) after confirming the distributions are normal. An additional preregistered hypothesis is reported in Appendix F. Results We calculated the semantic similarity between each of the 311 bathroom law Opponents’ justifications and all 2,655 HVBA premises (181 annotated for the two target values; 2,474 annotated for all other values) and find support for H4a. As Fig. 4a indicates, similarity to the two target values justifications ( M = 0.11, SD = 0.04) was greater on average than similarity to other annotations ( M = 0.08, SD = 0.03). The difference is significant by the standard of our preregistered test ( t (310) = 23.1, p < 0.001). We also find support for H4b. Opponents’ justifications are significantly more similar to premises invoking the two target values ( M = 0.11, SD = 0.04) than are Endorsers ( M = 0.10, SD = 0.04), confirming our second hypothesis ( t (599.9) = 1.8, p = 0.039). In a supplemental analysis (not preregistered), we found that the three most com- mon “best match” values for Opponents’ justifications were benchmarks in the HVBA annotated with the target value of Self-Direction (Action) (31%), followed by Humil- ity (20%), and the second target value of Universalism (Concern) (11%). We did not include Humility in our preregistered set of target value-categories, but it includes placing value on accepting life as it, which may be consistent with socially tolerant policy preferences. As shown in Fig. 4b, these three values accounted for over 60% of best matches for Opponents. While these values were also common best matches for Endorsers, the majority of their matches came from other values (e.g., Personal Secu- rity, 18%). Figure 4c provides several illustrations of best matches for open-ended bathroom law justifications. These illustrations suggest that the semantic similarity measure cannot distinguish between instances in which Endorsers advocate for val- ues on their own behalf (e.g., Item ID 235 in Fig. 4c) or on behalf of others. 23 We did not take a direct classification approach using the annotated data because best-in-class perfor- mance on classification tasks with many categories (20 in this case) is relatively low (Kiesel et al. 2023). 1 3 Political Behavior Fig. 4 Semantic similarity between preference justifications and value-annotated justifications. Note: (a) Distribution of semantic similarity to the two target values (left column) and all other values (right column) for each Opponent justification. (b) Proportions of best matches by group by value. (c) Illus- trative examples of best matches for Opponents and Endorsers Nevertheless, these text-based findings provide suggestive evidence that personal values act as a coherent source of intrinsic motivation for inclusionary preferences. The value of autonomy of action in particular appears to be important in the case of bathroom laws. While personal values may provide the intrinsic motivation required by the self-regulation activity observed in our fMRI study, further research is needed to link these processes more tightly. We note that the subject samples (42 people in Massachusetts; 602 people in Nebraska) also differ in a number of ways (Table F1), which limits our ability to draw firm conclusions about one study from the other. Discussion Across our two studies, we find evidence that inclusionary preferences are related to a value-driven thought process. We find nothing as coherent for exclusionary prefer- ences. In Study 1, we found evidence in bathroom law Opponents for the cognitive mechanism of self-regulation, previously associated with other forms of tolerance 1 3 Political Behavior in the social cognitive neuroscience literature (Amodio and Cikara 2021). Self-reg- ulation is a process that requires motivation and we found suggestive evidence in dmPFC that Opponents might be considering their personal values during the task. Study 1 thus suggested that value-related motivation might be an important correlate of inclusionary preferences. Study 2 used open-ended written justifications of bath- room law preferences to probe this possibility. Consistent with a value-driven differ- ence, we found that a small set of value-categories, autonomy of action in particular, characterized Opponents’ expressed logic to a greater extent than Endorsers. While it is not surprising that personal values are associated with policy pref- erences, it is important to recall that public support for policies is uneven where transgender people are concerned. Stated support for anti-discrimination measures is much higher than is support for access to bathrooms based on gender identity in a number of countries, including the United States (Flores et al. 2016). Our findings do not simply reinforce the importance of “values” writ large, but rather highlight how particular values might connect to specific policy preferences. These findings provide a new way to interpret the successes and failures of inter- ventions on bathroom law preferences reported by Broockman and Kalla (2016), Kalla and Broockman (2020), and Flores et al. (2021). All interventions exposed par- ticipants to the viewpoints of transgender people. And, across these studies, in-person interventions durably influenced policy preferences. But treatment effects weakened as the intensity of social contact involved in delivering the treatment decreased. Phone conversations had a weaker influence on policy preferences than in-person contact (Kalla and Broockman 2020) and online perspective-exposure treatments had null main effects (Flores et al. 2021). Our findings suggest a possible explana- tion for this pattern: interventions delivered interpersonally (face-to-face, over the phone) generate extrinsic motivation to engage in self-regulation, replacing the role of intrinsic motivation due to personal values already at work for those holding inclu- sionary preferences. Most interventions delivered online are unlikely to generate extrinsically motivated self-regulation because they do not involve real-time social interaction. Our theorized effect of social pressure is also consistent with the notion that intergroup conflict interventions can influence perceived social norms without shifting personal beliefs (e.g., Paluck 2009). Future intervention-based research into exclusionary policy preferences should therefore explore a self-regulation treatment component (e.g., Halperin et al. 2013; Hurtado-Parrado et al. 2019; Westerlund et al. 2020) and explicitly test the importance of interpersonal encouragement for treat- ment effectiveness. Limitations As with any research design, ours has limitations. Generalizing from a single neu- roimaging study is common practice but not ideal. Few, if any, similar studies into specific policy preferences exist, which means we cannot (yet) look for concordant findings to demonstrate generalizability. Our use of two demographically distinguish- able samples from Massachusetts and Nebraska ensures some diversity in the pool of participants on which our findings are based. Nevertheless, we cannot guarantee that 1 3 Political Behavior the cognitive mechanisms associated with bathroom law preferences in one sample necessarily hold in the other. Many potential sources of heterogeneity also could not be captured or tested by our design. Our studies were also correlational by necessity. Given the policy in which we were interested, we relied on pre-existing preferences to categorize participants in our studies. Without the ability to randomly assign group membership, we could not demonstrate that particular patterns of cognition or values caused preferences, or vice versa. Future research into the causal ordering of cognitive mechanisms and prefer- ences is still needed. Conclusion Our research considers an important political development over the last decade: the significant increase in efforts to curtail the rights of, and resources available to, transgender people. These efforts have resulted in declines in mental health (Lee et al. 2024) and prompted people to consider moving to different locales (James et al. 2024). The pursuit of restrictions aimed at transgender people cannot be explained entirely on the basis of attitudes (e.g., transphobia), which prompted our search for a better understanding of the cognitive mechanisms underlying inclusionary and exclu- sionary preferences. Research into the mental processes underlying exclusionary preferences has tended to focus on specific threat-related emotions, particularly disgust (e.g., Aarøe et al. 2017; Vanaman and Chapman 2020). We find no evidence, however, that emotional responses to imagined interactions with transgender people are suf- ficiently similar to a disgust response (or a fear response) to justify such a signifi- cant a focus on this mechanism. Nor do we find that emotional responses during these interactions distinguish bathroom law Opponents from Endorsers. We believe that the cognitive mechanism reconciling our findings with the broader emotions literature is self-regulation. Because most studies rely on self-reports to measure specific emotional responses and trait sensitivities (e.g., Miller et al. 2017; Vana- man and Chapman 2020), research designs have not been able to detect intermediate processes that affect end-state self-reporting. For example, self-reports of emotional end-states yield observably equivalent results when an emotion is absent and when it is down-regulated. Thus, measurement techniques may have masked the role of self- regulation in prior research. We also found no evidence of stronger mentalizing activity supporting inclusion- ary preferences, as interventions designed to influence preferences on bathroom laws (among other policies) have posited. We put forward an alternative explanation for the pattern of effects observed across these interventions. We argued that the self-reg- ulation we observed in bathroom law Opponents during our fMRI study could be the result of intrinsic motivation arising from particular personal values. The implication is that interventions seeking to encourage inclusionary preferences might be provid- ing extrinsic motivation to self-regulate via social pressure. This pathway stands in contrast to explanations of intervention treatment effects that credit the tasks directly (i.e., successfully fostering empathy). 1 3 Political Behavior In sum, our studies pave the way for both theoretical and practical advances in the study of inclusionary and exclusionary social policies and their sources of support. We argue that a value-driven account of specific policy preferences is consistent with the observation that support for transgender-related policies is uneven within individ- uals and not entirely correlated with intergroup attitudes. Examining the sources of support for specific policies offers insight into these more granular patterns of inclu- sionary and exclusionary preferences. Considering the toll that each exclusionary policy takes on a targeted community, these preferences are important to understand. Supplementary Information The online version contains supplementary material available at h t t ps : / / do i . o rg / 1 0 . 1 00 7 /s 1 1 1 0 9 -0 2 5 - 10 0 9 1 - x. Acknowledgements The authors thank David Broockman, Gabe Lenz, participants at the 2022 Annual Meeting of the International Society of Political Psychology (ISPP), attendees at the Research Workshop on American Politics at UC Berkeley, and three anonymous reviewers for their helpful comments and suggestions. Funding Partial financial support was received from the MIT School of Humanities, Arts, and Social Sci- ences (MLW), the Charles Koch Foundation (MLW), Beyond Conflict (MLW), the Patrick J. McGovern Foundation (RS), and the Guggenheim Foundation (RS). Data Availability All data and code required to replicate the analyses and figures are available from the Political Behavior Dataverse: https://doi.org/10.7910/DVN/4ETFPS. Our group-level fMRI data can be downloaded from Neurovault: h t tp s : / /i d e n t i fi er s . o rg / n e u r ov a ul t . c o l le c t i o n :1 3 0 2 2. Access to the raw s ur v e y data used in Study 2 must be requested from the Nebraska Bureau of Sociological Research. Declarations Compliance with Ethical Standards and IRB Approval All participants gave informed consent to partici- pate and have their results published. MIT’s Committee on the Use of Humans as Experimental Subjects (COUHES) approved the fMRI study. The text study was deemed exempt since the authors analyzed anonymized data previously collected by other researchers. Competing Interests The authors have no financial or non-financial interests to declare. Preregistration Both studies were preregistered on the Open Science Framework. Study 1 (fMRI): h t t p s: / / o sf . i o /b 8 d 5 h/ . Study 2 (text): https://osf.io/m83n7/. All deviations are indicated in the text. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit h t tp : / / cr e a t i v ec o mm o n s . o rg / l i ce n s e s/ b y / 4 . 0 /. 1 3 Political Behavior References Aarøe, Lene, Michael Bang Petersen, and Kevin Arceneaux. 2017. The behavioral immune system shapes political intuitions: Why and how individual differences in disgust sensitivity underlie opposition to immigration. American Political Science Review 111 (2): 277–294. h t t p s : // d o i .o r g / 10 . 1 0 17 / S 0 0 0 30 5 5 4 1 6 00 0 7 7 0. Ahn, Woo-Young., Kenneth T. Kishida, Xiaosi Gu, Terry Lohrenz, Ann Harvey, John R. Alford, Kevin B. Smith, et al. 2014. Nonpolitical images evoke neural predictors of political ideology. Current Biology 24 (22): 2693–2699. https://doi.org/10.1016/j.cub.2014.09.050. Amodio, David M., and Mina Cikara. 2021. The social neuroscience of prejudice. Annual Review of Psy- chology 72 (1): 439–469. h t tp s : / /d o i . or g / 1 0 . 11 4 6 / an n u r e v -p s y c h - 01 0 4 1 9- 0 5 0 92 8. Axt, Jordan R., Morgan A. Conway, Erin C. Westgate, and Nicholas R. Buttrick. 2021. Implicit transgen- der attitudes independently predict beliefs about gender and transgender people. Personality and Social Psychology Bulletin 47 (2): 257–274. https://doi.org/10.1177/0146167220921065. Bakker, Bert N., and Gijs Schumacher. 2025. Using Measures of Psychophysiological and Neural Activ- ity to Advance Understanding of Psychological Processes in Politics. In Handbook of Innovations in Political Psychology, ed. Ethan C. Busby, Christopher F. Karpowitz, and Cara J. Wong, 415–52. Northampton, MA: Edward Elgar Publishing. Bakker, Bert N., Gijs Schumacher, Claire Gothreau, and Kevin Arceneaux. 2020. Conservatives and liber- als have similar physiological responses to threats. Nature Human Behaviour 4 (6): 613–621. h t t p s: / / d oi . o r g/ 1 0 . 10 3 8 / s 4 15 6 2 - 0 2 0- 0 8 2 3- z. Barnett, Brian S., Ariana E. Nesbit, Renée M. Sorrentino. 2018. The transgender bathroom debate at the intersection of Politics, Law, Ethics, and science. The Journal of the American Academy of Psychia- try and the Law 46 (2): 232–241. https://doi.org/10.29158/JAAPL.003761-18 Berkman, Elliot T., Jordan L. Livingston, and Lauren E. Kahn. 2017. Finding the ‘Self’ in Self-Regulation: the Identity-Value model. Psychological Inquiry 28 (2–3): 77–98. ht t p s : / /d o i . or g / 1 0. 1 0 8 0/ 1 0 4 7 8 40 X . 2 0 1 7. 1 3 2 34 6 3. Bobo, Lawrence. 1983. Whites’ opposition to busing: Symbolic racism or realistic group conflict? Journal of Personality and Social Psychology 45 (6): 1196–1210. https://doi.org/10.1037/0022-3514.45.6.1196. Broockman, David, and Joshua Kalla. 2016. Durably reducing transphobia: A field experiment on door-to- door canvassing. Science 352 (6282): 220–224. https://doi.org/10.1126/science.aad9713. Brooks, Libby and Peter Walker. 2025. Trans People Banned from Toilets of Gender They Identify with, Says UK Minister. The Guardian: Society, April 27, 2025. h t tp s : / /w w w . t h eg u a r di a n . c o m/ s o c i e ty / 2 0 2 5 / a p r/ 2 7 / tr a n s - p eo p l e - b an n e d -f r o m - t oi l e t s - of - g e nd e r - t h ey - i d e n ti f y - wi t h - s a ys - u k -m i n i st e r Bruneau, Emile G., Nicholas Dufour, and Rebecca Saxe. 2012. Social cognition in members of conflict groups: Behavioural and neural responses in Arabs, Israelis and South Americans to each other’s misfortunes. Philosophical Transactions of the Royal Society B: Biological Sciences 367 (1589): 717–730. https://doi.org/10.1098/rstb.2011.0293. Bureau of Sociological Research. 2018. 2018 Nebraska Annual Social Indicators Survey. Casado-Aranda, Luis-Alberto, Juan Sánchez-Fernández, and Teodoro Luque-Martínez. 2022. Modulating the neural bases of political communications: Political involvement and perception of the economic situation. Political Behavior 44 (3): 1247–1270. https://doi.org/10.1007/s11109-020-09657-8 Chakrabarti, Parijat, and Margaret Frye. 2017. A mixed-methods framework for analyzing text data: Inte- grating computational techniques with qualitative methods in demography. Demographic Research 37: 1351–1382. Clifford, Scott, and Spencer Piston. 2017. Explaining public support for counterproductive homelessness policy: The role of disgust. Political Behavior 39 (2): 503–525. h tt p s : / / do i . o r g /1 0 . 1 00 7 / s 1 1 10 9 - 0 1 6 -9 3 6 6 -4 . Cottrell, Catherine A., and Steven L. Neuberg. 2005. Different emotional reactions to different groups: A sociofunctional threat-based approach to “prejudice.” Journal of Personality and Social Psychology 88 (5): 770–789. https://doi.org/10.1037/0022-3514.88.5.770. Cox, Jennifer, Marissa Stanziani, C. Adam Coffey, Elizabeth Bownes, Stephanie Brooks Holliday, Lauren T. Meaux. 2022. Your rights end where mine begin’: A Mixed-Methods study of moral foundations theory and support for bathroom bills. Sexuality Research and Social Policy 19(2):453–469. h tt p s : / / do i . o r g /1 0 . 1 00 7 / s 1 3 17 8 - 0 2 1 -0 0 5 5 9- w 1 3 Political Behavior Dodell-Feder, David, Jorie Koster-Hale, Marina Bedny, and Rebecca Saxe. 2011. fMRI item analysis in a theory of mind task. Neuroimage 55 (2): 705–712. h tt p s :/ / do i . or g / 10 . 1 0 16 / j. n e u ro i m ag e . 20 1 0 . 12 . 04 0. Dufour, Nicholas, Elizabeth Redcay, Liane Young, Penelope L. Mavros, Joseph M. Moran, Christina Tri- antafyllou, John D. E. Gabrieli, and Rebecca Saxe. 2013. Similar brain activation during false belief tasks in a large sample of adults with and without autism. PLoS One 8 (9): e75468. h tt p s :/ / d o i. o rg / 1 0 .1 3 7 1/ j o ur n a l .p o ne . 0 0 75 4 6 8. Eickhoff, Simon B., B. T. Thomas Yeo, and Sarah Genon. 2018. Imaging-based parcellations of the human brain. Nature Reviews Neuroscience 19 (11): 672–686. https://doi.org/10.1038/s41583-018-0071-7 Eickhoff, Simon B., Angela R. Laird, Christian Grefkes, Ling E. Wang, Karl Zilles, and Peter T. Fox. 2009. Coordinate-based activation likelihood estimation meta-analysis of neuroimaging data: A random- effects approach based on empirical estimates of spatial uncertainty. Human Brain Mapping 30 (9): 2907–2926. https://doi.org/10.1002/hbm.20718 Federico, Christopher M., and Jim Sidanius. 2002. Sophistication and the antecedents of whites’ racial policy attitudes: racism, ideology, and affirmative action in America. Public Opinion Quarterly 66 (2): 145–176. https://doi.org/10.1086/339848 Fedorenko, Evelina, John Duncan, and Nancy Kanwisher. 2013. Broad domain generality in focal regions of frontal and parietal cortex. Proceedings of the National Academy of Sciences of the United States of America 110 (41): 16616–21. https://doi.org/10.1073/pnas.1315235110. Fehlbaum, Lynn V., Réka. Borbás, Katharina Paul, Simon B. Eickhoff, and Nora M. Raschle. 2022. Early and late neural correlates of mentalizing: ALE meta-analyses in adults, children and adolescents. Social Cognitive and Affective Neuroscience 17 (4): 351–366. https://doi.org/10.1093/scan/nsab105. Flores, Andrew R., Taylor N. T. Brown, and Andrew S. Park. 2016. Public Support for Transgender Rights: A Twenty-Three Country Survey. Williams Institute, UCLA School of Law. ht t p s :/ / w il l i am s i n st i tu t e .l a w .u c la . e d u/ p u bl i c at i o n s/ t ra n s - ri g h ts - 2 3- c o u nt r y- s u r ve y / Flores, Andrew R., Donald P. Haider-Markel, Daniel C. Lewis, Patrick R. Miller, Barry L. Tadlock, and Jami K. Taylor. 2018a. Challenged expectations: Mere exposure effects on attitudes about trans- gender people and rights. Political Psychology 39 (1): 197–216. https://doi.org/10.1111/pops.12402. Flores, Andrew R., Donald P. Haider-Markel, Daniel C. Lewis, Patrick R. Miller, Barry L. Tadlock, and Jami K. Taylor. 2018b. Transgender prejudice reduction and opinions on transgender rights: Results from a mediation analysis on experimental data. Research & Politics 5 (1): 2053168018764945. https://doi.org/10.1177/2053168018764945 Flores, Andrew R., Donald P. Haider-Markel, Daniel C. Lewis, Patrick R. Miller, and Jami K. Taylor. 2021. Antidiscrimination interventions, political ads on transgender rights, and public opinion: Results from two survey experiments on adults in the United States. Frontiers in Psychology. h t t ps : / /d o i .o r g / 10 . 33 8 9 / fp s y g. 2 0 21 . 7 2 93 2 2. Freeder, Sean, Gabriel S. Lenz, and Shad Turney. 2019. The importance of knowing ‘What goes with what’: Reinterpreting the evidence on policy attitude stability. The Journal of Politics 81 (1): 274– 290. https://doi.org/10.1086/700005. Frith, Chris D., and Uta Frith. 2006. The neural basis of mentalizing. Neuron 50 (4): 531–534. h tt p s :/ / d o i . o rg / 1 0 .1 0 1 6/ j . ne u r o n. 2 00 6 . 05 . 0 0 1. GLAAD. 2022. GLAAD Media Reference Guide. 11th ed. https://glaad.org/reference/ Gross, James J. 2015. Emotion regulation: Current status and future prospects. Psychological Inquiry 26 (1): 1–26. https://doi.org/10.1080/1047840X.2014.940781. Haas, Ingrid J. 2016. Political Neuroscience. In Neuroimaging Personality, Social Cognition, and Charac- ter, ed. John R. Absher and Jasmin Cloutier, 355–70. San Diego: Academic Press. ht t p s :/ / d oi . o rg / 1 0 .1 0 16 / B 9 78 - 0 -1 2 - 80 0 93 5 - 2. 0 0 01 9 - 1. Haas, Ingrid J., Melissa N. Baker, and Frank J. Gonzalez. 2021. Political uncertainty moderates neural evaluation of incongruent policy positions. Philosophical Transactions of the Royal Society B: Bio- logical Sciences 376(1822):20200138. https://doi.org/10.1098/rstb.2020.0138 Hainmueller, Jens, and Daniel J. Hopkins. 2014. Public attitudes toward immigration. Annual Review of Political Science 17 (1): 225–249. h t t p s: / /d o i . or g / 10 . 1 14 6 / a nn u re v - po l i sc i -1 0 2 5 12 - 1 94 8 1 8. Hallenbeck, Erin, Susan Dicklitch-Nelson, Marin Moore, Amy R. Moreno, and Juliana Stoll. 2025. Global LGBT Rights from 2011–2022: A Mixed Record. F&M Global Barometers. h t t p s: / /w w w . fa n d mg l o b a l b a ro m et e r s .o r g /w p - co n t e nt / up l o a ds / 2 02 5 / 05 / FI N A L2 0 2 5- F M G B- A nn u a l -R e p or t . pd f Halperin, Eran, Roni Porat, Maya Tamir, and James J. Gross. 2013. Can emotion regulation change politi- cal attitudes in intractable conflicts? From the laboratory to the field. Psychological Science 24 (1): 106–111. https://doi.org/10.1177/0956797612452572. 1 3 Political Behavior Hasenbush, Amira, Andrew R. Flores, and Jody L. Herman. 2019. Gender identity nondiscrimination laws in public accommodations: A review of evidence regarding safety and privacy in public restrooms, locker rooms, and changing rooms. Sexuality Research & Social Policy 16 (1): 70–83. ht t p s :/ / d o i. o r g / 1 0 .1 0 0 7 /s 1 3 1 78 - 0 1 8- 0 3 3 5- z Haxby, James V., M. Ida. Gobbini, Maura L. Furey, Alumit Ishai, Jennifer L. Schouten, and Pietro Pietrini. 2001. Distributed and overlapping representations of faces and objects in ventral temporal cortex. Science 293 (5539): 2425–2430. https://doi.org/10.1126/science.1063736 Hopkins, Daniel J., John Sides, and Jack Citrin. 2019. The muted consequences of correct information about immigration. The Journal of Politics 81 (1): 315–320. https://doi.org/10.1086/699914. Huettel, S.A., A.W. Song, and G. McCarthy. 2004. Functional magnetic resonance imaging. Sunderland MA: Sinauer Associates. Hurtado-Parrado, Camilo, Myriam Sierra-Puentes, Mohammed El Hazzouri, Alexandra Morales, Diana Gutiérrez-Villamarín, Laura Velásquez, Andrea Correa-Chica, et al. 2019. Emotion regulation and attitudes toward conflict in Colombia: Effects of reappraisal training on negative emotions and sup- port for conciliatory and aggressive statements. Frontiers in Psychology 10. h tt p s : // d o i .o r g / 10 . 3 3 89 / f p sy g . 2 01 9 . 0 09 0 8 . Hutzler, Florian. 2014. Reverse inference is not a fallacy per se: Cognitive processes can be inferred from functional imaging data. NeuroImage 84 (January): 1061–69. h t t p s: / / d oi . or g / 1 0. 1 0 1 6/ j . n eu r o i ma g e . 20 1 2 . 12 . 0 7 5. Inzlicht, Michael, Kaitlyn M. Werner, Julia L. Briskin, and Brent W. Roberts. 2021. Integrating models of self-regulation. Annual Review of Psychology 72 (1): 319–345. h t tp s : / /d o i . or g / 1 0. 1 1 4 6/ a n n ur e v - ps y c h -0 6 10 2 0 - 10 5 7 2 1. Jacoby, Nir, Marika Landau-Wells, Jacob Pearl, Alexandra Paul, Emily B. Falk, Emile G. Bruneau, and Kevin N. Ochsner. 2024. Partisans process policy-based and identity-based messages using disso- ciable neural systems. Cerebral Cortex 34 (9): bhae368. https://doi.org/10.1093/cercor/bhae368. James, Sandy E., Jody L. Herman, Susan Rankin, Mara Keisling, Lisa Mottet, and Ma’ayan Anafi. 2016. The Report of the 2015 U.S. Transgender Survey. Washington, D.C.: National Center for Transgen- der Equality. h t t ps : / / tr a n s eq u a l it y . o rg / s i te s / d ef a u l t/ fi l e s / do c s / us t s / US T S -F u l l- R e p or t - D ec 1 7 . pd f James, Sandy E., Jody L. Herman, Laura E. Durso, and Rodrigo Heng-Lehtinen. 2024. Early Insights: A Report of the 2022 U.S. Transgender Survey. Washington, D.C.: National Center for Transgender Equality. h t t ps : / / tr a n s eq u a l it y . o rg / s i te s / d ef a u lt / fi l e s/ 2 0 2 4- 0 2 / 20 2 2 % 20 U S T S% 2 0 E ar l y % 20 I n s ig h t s %2 0 R e po r t _ FI N A L .p d f Jones, Philip Edward, and Paul R. Brewer. 2020. Elite cues and public polarization on transgender rights. Politics, Groups, and Identities 8 (1): 71–85. ht t p s :/ / d o i. o r g /1 0 . 10 8 0 /2 1 5 6 55 0 3 . 20 1 8 . 14 4 1 7 22 . Jost, John T., H. Hannah Nam, David M. Amodio, and Jay J. Van Bavel. 2014. Political neuroscience: The beginning of a beautiful friendship. Political Psychology 35 (S1): 3–42. ht t p s :/ / d o i. o r g /1 0 . 1 11 1 / p o p s . 1 21 6 2 . Kalla, Joshua L., and David E. Broockman. 2020. Reducing exclusionary attitudes through interpersonal conversation: Evidence from three field experiments. American Political Science Review 114 (2): 410–425. https://doi.org/10.1017/S0003055419000923. Kalla, Joshua L., and David E. Broockman. 2023. Which narrative strategies durably reduce prejudice? Evidence from field and survey experiments supporting the efficacy of perspective-getting. American Journal of Political Science 67 (1): 185–204. https://doi.org/10.1111/ajps.12657. Kam, Cindy D., and Beth A. Estes. 2016. Disgust sensitivity and public demand for protection. The Jour- nal of Politics 78 (2): 481–496. https://doi.org/10.1086/684611. Kanwisher, Nancy. 2010. Functional specificity in the human brain: A window into the functional architec- ture of the mind. Proceedings of the National Academy of Sciences of the United States of America 107 (25): 11163–70. https://doi.org/10.1073/pnas.1005062107. Kazyak, Emily, Kelsy Burke, Rosalind Kichler, and Lora McGraw. 2021. Pee in peace’ or ‘Make everyone uncomfortable’: Public perceptions of transgender rights. Socius 7 (January): 23780231211055541. https://doi.org/10.1177/23780231211055541. Kiesel, Johannes, Milad Alshomary, Nicolas Handke, Xiaoni Cai, Henning Wachsmuth, and Benno Stein. 2022. Identifying the Human Values Behind Arguments. In Proceedings of the 60th Annual Meet- ing of the Association for Computational Linguistics (Volume 1: Long Papers), 4459–71. Dublin, Ireland: Association for Computational Linguistics. h tt p s : // d o i .o r g / 10 . 1 8 65 3 / v 1/ 2 0 2 2. a c l -l o n g .3 0 6 1 3 Political Behavior Kiesel, Johannes, Milad Alshomary, Nailia Mirzakhmedova, Maximilian Heinrich, Nicolas Handke, Hen- ning Wachsmuth, and Benno Stein. 2023. SemEval-2023 Task 4: ValueEval: Identification of Human Values Behind Arguments. In Proceedings of the 17th International Workshop on Semantic Evalua- tion (SemEval-2023), edited by Atul Kr. Ojha, A. Seza Doğruöz, Giovanni Da San Martino, Harish Tayyar Madabushi, Ritesh Kumar, and Elisa Sartori, 2287–2303. Toronto, Canada: Association for Computational Linguistics. h t tp s :/ / d oi . o rg / 1 0 .1 8 65 3 / v1 / 2 02 3 .s e m e va l - 1. 3 1 3 Kinder, Donald R., and David O. Sears. 1981. Prejudice and politics: Symbolic racism versus racial threats to the good life. Journal of Personality and Social Psychology 40 (3): 414–431. h t tp s : / /d o i. o r g/ 1 0 .1 0 37 / 0 0 22 - 3 51 4 . 40 . 3. 4 1 4. Kragel, Philip A., and Kevin S. LaBar. 2016. Decoding the nature of emotion in the brain. Trends in Cogni- tive Sciences 20 (6): 444–455. https://doi.org/10.1016/j.tics.2016.03.011. Kriegeskorte, Nikolaus, Marieke Mur, and Peter Bandettini. 2008. Representational similarity analysis – Connecting the branches of systems neuroscience. Frontiers in Systems Neuroscience 2 (November). https://doi.org/10.3389/neuro.06.004.2008 Krupnikov, Yanna, Spencer Piston, and Nichole M. Bauer. 2016. Saving face: Identifying voter responses to Black candidates and female candidates. Political Psychology 37 (2): 253–273. h t t p s: / /d o i .o r g /1 0 .1 1 1 1 /p o p s. 1 2 26 1. Kushner Gadarian, Shana, and Ted Brader. 2023. Emotion and Political Psychology. In The Oxford Hand- book of Political Psychology, edited by Leonie Huddy, David O. Sears, Jack S. Levy, and Jennifer Jerit, 191–247. Oxford University Press. ht t ps : / /d o i .o r g/ 1 0 . 10 9 3 /o x f or d hb / 9 78 0 1 97 5 4 1 30 2 .0 1 3 .5 Landau-Wells, Marika. 2018. Dealing with Danger: Threat Perception and Policy Preferences. PhD thesis, Cambridge, M.A.: MIT. http://hdl.handle.net/1721.1/118222 Landau-Wells, Marika, and Rebecca Saxe. 2020. Political preferences and threat perception: Opportunities for neuroimaging and developmental research. Current Opinion in Behavioral Sciences 34 (August): 58–63. https://doi.org/10.1016/j.cobeha.2019.12.002 Landau-Wells, Marika. 2024. Building from the brain: Advancing the study of threat perception in inter- national relations. International Organization 78 (4): 627–667. h tt p s: / / do i . or g /1 0 . 1 01 7 / S0 0 2 08 1 83 2 4 00 0 3 28 . Lee, Wilson Y., J. Nicholas Hobbs, Steven Hobaica, Jonah P. DeChants, Myeshia N. Price, and Ronita Nath. 2024. State-level anti-transgender laws increase past-year suicide attempts among transgender and non-binary young people in the USA. Nature Human Behaviour 8 (11): 2096–2106. h t tp s : // d o i . o r g/ 1 0 .1 0 3 8/ s 41 5 6 2 -0 2 4 -0 1 9 79 - 5. Lenz, Gabriel S. 2013. Follow the leader? How voters respond to politicians’ policies and performance. Chicago, IL: University of Chicago Press. Lewis, Tené T., Courtney D. Cogburn, and David R. Williams. 2015. Self-reported experiences of discrim- ination and health: Scientific advances, ongoing controversies, and emerging issues. Annual Review of Clinical Psychology 11: 407–440. h t t ps : // d o i .o r g /1 0 . 11 4 6/ a n nu r e v- c l i np s y- 0 3 2 81 4 - 11 2 7 28 . Lewis, Daniel C., Andrew R. Flores, Donald P., Haider-Markel, Patrick R. Miller, and Jami K. Taylor. 2022. Transitioning opinion? Assessing the dynamics of public attitudes toward transgender rights. Public Opinion Quarterly 86 (2):343–368. https://doi.org/10.1093/poq/nfac014 Lewis, Daniel C., Andrew R. Flores, Donald P. Haider-Markel, Patrick R. Miller, and Jami K. Taylor. 2023. American Attitudes Toward Transgender Rights. The Gender Policy Report. h t t ps : / /g e n d er p ol i c yr e p or t .u m n . ed u / am e r ic a n- a t ti t u de s - t ow a rd - t ra n s ge n de r - r ig h t s/ Lewis, Daniel C., Andrew R. Flores, Donald P. Haider-Markel, Patrick R. Miller, and Jami K. Taylor. 2024. Cultural threat, outgroup discrimination, and attitudes toward transgender rights. Political Behavior 46 (4): 2401–2426. https://doi.org/10.1007/s11109-024-09924-y. Lieberman, Matthew D., Mark A. Straccia, Meghan L. Meyer, Meng Du, and Kevin M. Tan. 2019. Social, Self, (Situational), and affective processes in medial prefrontal cortex (MPFC): Causal, Multivariate, and reverse inference evidence. Neuroscience & Biobehavioral Reviews 99 (April): 311–328. h t tp s : / / d oi . o r g/ 1 0 .1 0 1 6/ j .n e u bi o r ev . 2 0 18 . 12 . 0 21 Luhur, Winston, Taylor N. T. Brown, and Andrew Flores. 2019. Public Opinion of Transgender Rights in the United States. Los Angeles: Williams Institute, UCLA School of Law. h tt p s: / / w il l i am s i ns t it u t e. l a w. u c l a. e du / p ub l i ca t io n s / pu b l ic - o pi n io n - tr a n s- r i g ht s -u s / Mattan, Bradley D., Jennifer T. Kubota, Tianyi Li, Tzipporah P. Dang, and Jasmin Cloutier. 2018. Moti- vation modulates brain networks in response to faces varying in race and status: A multivariate approach. eNeuro. https://doi.org/10.1523/ENEURO.0039-18.2018 McDermott, Rose. 2004. The feeling of rationality: The meaning of neuroscientific advances for political science. Perspectives on Politics 2 (4): 691–706. https://doi.org/10.1017/S1537592704040459. 1 3 Political Behavior McGraw, Kathleen M. 2000. Contributions of the cognitive approach to political psychology. Political Psychology 21 (4): 805–832. https://doi.org/10.1111/0162-895X.00217. McLaren, Lauren. 2003. Anti-immigrant prejudice in Europe: Contact, threat perception, and preferences for the exclusion of migrants. Social Forces 81 (3): 909–936. https://doi.org/10.1353/sof.2003.0038. McRae, Kateri, Soo Hyun Rhee, Justine M. Gatt, Detre Godinez, Leanne M. Williams, and James J. Gross. 2017. Genetic and environmental influences on emotion regulation: A twin study of cognitive reap- praisal and expressive suppression. Emotion 17 (5): 772–777. https://doi.org/10.1037/emo0000300. Michelson, Melissa R., and Brian F. Harrison. 2020. Transforming prejudice: Identity, fear, and transgen- der rights. Oxford University Press. Mikhaylov, Slava, Michael Laver, and Kenneth Benoit. 2012. Coder reliability and misclassification in the human coding of party manifestos. Political Analysis 20 (1): 78–91. h t tp s :/ / d oi . o rg / 10 . 1 09 3 / pa n /m p r 04 7. Miller, Patrick R., Andrew R. Flores, Donald P. Haider-Markel, Daniel C. Lewis, Barry L. Tadlock, and Jami K. Taylor. 2017. Transgender politics as body politics: Effects of disgust sensitivity and authori- tarianism on transgender rights attitudes. Politics, Groups, and Identities 5 (1): 4–24. h tt p s: / / do i . or g /1 0 . 10 8 0 /2 1 56 5 50 3 .2 0 1 6. 1 2 60 4 82 . Moore, Adam, Sujin Hong, and Laura Cram. 2021. Trust in information, political identity and the brain: An interdisciplinary fMRI study. Philosophical Transactions of the Royal Society B: Biological Sciences 376 (1822): 20200140. https://doi.org/10.1098/rstb.2020.0140. Movement Advancement Project. 2025. Equality Maps: Bans on Transgender People’s Use of Public Bath- rooms & Facilities According to Their Gender Identity. ht t p :/ / ww w . ma p r es e ar c h .o r g /e q ua l i ty m a ps / n o n d is c r im i na t i on / b at h ro o m _b a n s Muennighoff, Niklas, Nouamane Tazi, Loic Magne, and Nils Reimers. 2023. MTEB: Massive Text Embedding Benchmark. In Proceedings of the 17th Conference of the European Chapter of the Asso- ciation for Computational Linguistics, 2014–37. Dubrovnik, Croatia: Association for Computational Linguistics. https://aclanthology.org/2023.eacl-main.148 Nam, H. Hannah, John T. Jost, Lisa Kaggen, Daniel Campbell-Meiklejohn, and Jay J. Van Bavel. 2018. Amygdala structure and the tendency to regard the social system as legitimate and desirable. Nature Human Behaviour 2 (2): 133–138. https://doi.org/10.1038/s41562-017-0248-5 Niazi, Shuriah. 2023. Trans People in India Demand Equal Access to Public Toilets. Reuters, August 2, 2023. h t t ps : // w w w. r e ut e rs . c om / a rt i cl e / wo r l d/ u s/ f e at u re - tr a n s- p eo p l e- i n -i n di a - de m a nd - eq u a l- a c ce s s - t o -p u b li c -t o i le t s -i d US L 8 N3 7 S 4W X / Nummenmaa, Lauri. and Heini Saarimäki. 2019. Emotions as discrete patterns of systemic activity. Neu- roscience Letters, functional neuroimaging of the emotional Brain, 693 (February): 3–8. h t tp s :/ / do i . or g /1 0 . 10 1 6 /j . ne u l et . 2 01 7 .0 7 . 01 2 Ochsner, Kevin N., and James J. Gross. 2013. The Neural Basis of Emotion and Emotion Regulation: A Valuation Perspective. In Handbook of Emotion Regulation, ed. James J. Gross, 23–42. New York: Guilford Publications. Paluck, Elizabeth Levy. 2009. What’s in a norm? Sources and processes of norm change. Journal of Per- sonality and Social Psychology 96 (3): 594–600. https://doi.org/10.1037/a0014688. Paluck, Elizabeth Levy, Seth A. Green, and Donald P. Green. 2019. The contact hypothesis re-evaluated. Behavioural Public Policy 3 (2): 129–158. https://doi.org/10.1017/bpp.2018.25. Patterson, Charlotte J., Martín-José Sepúlveda, and Jordyn White. eds. 2020. Understanding the Well-Being of LGBTQI + Populations. Washington, D.C.: National Academies. https://doi.org/10.17226/25877 Pew Research Center. 2022. U.S. Adults Have Complex Views on Gender Identity and Transgender Issues. h t tp s : / / w ww . p ew r e s e a r c h . or g /s o c ia l - tr e nd s / 20 2 2 /0 6 /2 8 /a m er i c an s - co m pl e x - v i e ws - on - g en d e r - i de n t it y - an d -t r a ns g e nd e r- i s su e s / Pew Research Center. 2025. Americans Have Grown More Supportive of Restrictions for Trans People in Recent Years. Pew Research Center. h t t ps : // w w w. p e wr e se a r ch . o rg / sh o r t- r e ad s /2 0 25 / 02 / 2 6/ a m er i ca n s -h a v e- g ro w n -m o r e- s up p o rt i v e- o f- r e st r i ct i on s - fo r - tr a ns - p eo p l e- i n- r e ce n t -y e ar s / Pew Research Center. 2020. The Global Divide on Homosexuality Persists. Washington, D.C. h tt p s :/ / w w w . p e w r e s e a r c h . o r g / g l o b a l / w p - c o n t e n t / u p l o a d s / s i t e s / 2 / 2 0 2 0 / 0 6 / P G _ 2 0 2 0 . 0 6 . 2 5 _ G l o b a l - V i e w s - H o m o s e xu a li t y _F I N AL . pd f Pew Research Center. 2019. Majority of Public Favors Same-Sex Marriage, but Divisions Persist. Wash- ington, D.C. h tt p s :/ / w ww . pe w r es e a rc h .o r g /p o l it i cs / w p- c o nt e nt / u pl o a ds / si t e s/ 4 / 20 1 9/ 0 5 /2 0 1 9_ 0 5_ 1 4 -S a m e- S ex - M ar r i ag e -F I N AL . p df Poldrack, Russell A. 2011. Inferring mental states from neuroimaging data: From reverse inference to large-scale decoding. Neuron 72 (5): 692–697. https://doi.org/10.1016/j.neuron.2011.11.001. 1 3 Political Behavior Poldrack, Russell A., and Tal Yarkoni. 2016. From brain maps to cognitive ontologies: Informatics and the search for mental structure. Annual Review of Psychology 67 (January): 587–612. h t tp s : / /d o i . or g / 1 0. 1 1 46 / a nn u r e v- p s y ch - 1 2 24 1 4 - 03 3 7 2 9 Poldrack, Russell A., and Jeanette A. Mumford. 2009. Independence in ROI analysis: Where is the voodoo? Social Cognitive and Affective Neuroscience 4 (2): 208–213. https://doi.org/10.1093/scan/nsp011. Poldrack, Russell A., Jeanette A. Mumford, and Thomas E. Nichols. 2011. Handbook of functional MRI data analysis. Cambridge University Press. Reimers, Nils. 2022. Sentence-Transformers. https://www.sbert.net/ Reimers, Nils, and Iryna Gurevych. 2019. Sentence-BERT: Sentence Embeddings Using Siamese BERT- networks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP- IJCNLP), edited by Kentaro Inui, Jing Jiang, Vincent Ng, and Xiaojun Wan, 3982–92. Hong Kong, China: Association for Computational Linguistics. https://doi.org/10.18653/v1/D19-1410 Rubin, Mark. 2021. When to adjust alpha during multiple testing: A consideration of disjunction, conjunc- tion, and individual testing. Synthese 199 (3): 10969–11000. h t t p s: / /d o i . or g / 1 0. 1 0 0 7/ s 1 12 2 9 -0 2 1 - 03 2 7 6 -4 . Saxe, Rebecca, and Nancy Kanwisher. 2003. People thinking about thinking people: The role of the tem- poro-parietal junction in ‘Theory of mind. Neuroimage 19 (4): 1835–1842. h tt p s : // d o i .o r g / 10 . 10 1 6 / S1 0 5 3 -8 1 1 9 (0 3 ) 00 2 3 0- 1. Schreiber, Darren, Greg Fonzo, Alan Simmons, Chris Dawes, Taru Flagan, and Martin Paulus. 2022. Neu- ral nonpartisans. Journal of Elections Public Opinion and Parties 32 (3): 576–595. h tt p s: / / d oi . o r g/ 1 0 . 10 8 0 / 17 4 5 7 28 9 . 2 02 0 . 1 80 1 6 95 . Schwartz, Shalom H. 1994. Are there universal aspects in the structure and contents of human values? Journal of Social Issues 50 (4): 19–45. h t tp s : / /d o i. o r g /1 0 . 1 11 1 / j .1 5 4 0- 4 5 60 . 1 9 94 . t b 01 1 9 6 .x . Schwartz, Shalom H., Jan Cieciuch, Michele Vecchione, Eldad Davidov, Ronald Fischer, Constanze Beier- lein, Alice Ramos, et al. 2012. Refining the theory of basic individual values. Journal of Personality and Social Psychology 103 (4): 663–688. https://doi.org/10.1037/a0029393. Senholzi, Keith B., and Jennifer T. Kubota. 2016. The Neural Mechanisms of Prejudice Intervention. In Neuroimaging Personality, Social Cognition, and Character, ed. John R. Absher and Jasmin Clout- ier, 337–54. San Diego, CA: Elsevier Academic Press. h t t p s: / / d oi . o r g/ 1 0 . 10 1 6/ B 9 7 8- 0 - 1 2- 8 0 0 93 5 - 2 . 0 0 01 8 - X Shine, James M., and Russell A. Poldrack. 2018. Principles of dynamic network reconfiguration across diverse brain states. NeuroImage 180: 396–405. h t t ps : / / do i . o rg / 1 0 .1 0 16 / j . ne u r o im a g e .2 0 1 7 .0 8 . 0 10 . Silvers, Jennifer A., Tor D. Wager, Jochen Weber, and Kevin N. Ochsner. 2015. The neural bases of unin- structed negative emotion modulation. Social Cognitive and Affective Neuroscience 10 (1): 10–18. https://doi.org/10.1093/scan/nsu016. Silvers, Jennifer A., and João F. Guassi Moreira. 2019. Capacity and tendency: A neuroscientific frame- work for the study of emotion regulation. Neuroscience Letters Functional Neuroimaging of the Emotional Brain 693 (February): 35–39. https://doi.org/10.1016/j.neulet.2017.09.017 Tamir, Maya. 2016. Why do people regulate their emotions? A taxonomy of motives in emotion regula- tion. Personality and Social Psychology Review 20 (3): 199–222. h t t ps : / / do i . o rg / 1 0 .1 1 7 7 /1 0 8 8 86 8 3 1 55 8 6 3 25 . Teed, Adam R., Jelena Rakic, Daniel B. Mark, and Daniel C. Krawcyzk. 2020. Relative activation patterns associated with self-transcendent and self-enhancement core values: An fMRI study of basic human values theory concepts in males. Social Neuroscience 15 (1): 1–14. ht t p s :/ / do i . o rg / 1 0 .1 0 8 0 /1 7 4 70 9 1 9. 2 0 1 9. 1 5 9 88 9 3. TGEU. 2025. Trans Rights Index & Map 2025. h t tp s : / /t g e u .o r g / tr a n s- r i gh t s - in d e x -m a p - 20 2 5/ Thapa, Raksha, Edwin van Teijlingen, Pramod Raj Regmi, and Vanessa Heaslip. 2021. Caste exclusion and health discrimination in South Asia: A systematic review. Asia-Pacific Journal of Public Health 33 (8): 828–838. https://doi.org/10.1177/10105395211014648. Theodoridis, Alexander G., and Amy J. Nelson. 2012. Of BOLD claims and excessive fears: A call for caution and patience regarding political neuroscience. Political Psychology 33 (1): 27–43. h tt p s : // d o i .o r g / 10 . 11 1 1 / j. 1 4 6 7- 9 2 2 1. 2 0 1 1. 0 0 8 60 . x. Trans Legislation Tracker. 2025. 2025 Anti-Trans bills: Trans Legislation Tracker. ht t p s :/ / t ra n s le g i s la t i o n. c o m Van Bavel, Jay J., and Andrea Pereira. 2018. The partisan brain: An identity-based model of political belief. Trends in Cognitive Sciences 22 (3): 213–224. https://doi.org/10.1016/j.tics.2018.01.004. 1 3 Political Behavior van Leeuwen, Florian, Yoel Inbar, Michael Bang Petersen, Lene Aarøe, Pat Barclay, Fiona Kate Barlow, and Mícheál de Barra et al. 2023. Disgust sensitivity relates to attitudes toward gay men and lesbian women across 31 nations. Group Processes & Intergroup Relations 26 (3):629–651. h tt p s : // d o i .o r g/ 1 0 . 11 7 7 / 13 6 8 4 30 2 2 1 10 6 7 1 51 Vanaman, Matthew E., and Hanah A. Chapman. 2020. Disgust and disgust-driven moral concerns pre- dict support for restrictions on transgender bathroom access. Politics and the Life Sciences 39 (2): 200–214. https://doi.org/10.1017/pls.2020.20. Vasilopoulos, Pavlos, George E. Marcus, Nicholas A. Valentino, and Martial Foucault. 2019. Fear, anger, and voting for the far right: Evidence from the November 13, 2015 Paris terror attacks. Political Psychology 40 (4): 679–704. https://doi.org/10.1111/pops.12513. Webster, Steven W., and Bethany Albertson. 2022. Emotion and politics: Noncognitive psychological biases in public opinion. Annual Review of Political Science 25 (1): 401–418. h t t p s: / / d oi . o r g/ 1 0 . 11 4 6 / an n u r ev - p o li s ci - 0 5 11 2 0 - 10 5 3 5 3. Westerlund, Minja, Pekka Santtila, and Jan Antfolk. 2020. Regulating emotions under exposure to nega- tive out-group-related news material results in increased acceptance of out-groups. The Journal of Social Psychology 160 (3): 357–372. ht t p s :/ / d o i. o r g /1 0 . 1 08 0 / 0 02 2 45 4 5 . 20 1 9 . 16 7 5 5 75 . Yang, Xudong, Hongli Yan, Anguo Zhang, Pan Xu, Sio Hang Pan, Mang I. Vai, and Yueming Gao. 2024. Emotion recognition based on multimodal physiological signals using spiking feed-forward neural networks. Biomedical Signal Processing and Control 91(May):105921. h t t p s: / / d oi . o rg / 1 0. 1 0 1 6/ j . b s p c . 2 02 3 . 1 05 9 2 1 Yarkoni, Tal, Russell A. Poldrack, Thomas E. Nichols, David C. Van Essen, and Tor D. Wager. 2011. Large-scale automated synthesis of human functional neuroimaging data. Nature Methods 8 (8): 665–670. https://doi.org/10.1038/nmeth.1635. Zaller, John R. 1992. The nature and origins of mass opinion. Cambridge, U.K.: Cambridge University Press. Zhou, Feng, Weihua Zhao, Ziyu Qi, Yayuan Geng, Shuxia Yao, Keith M. Kendrick, Tor D. Wager, and Benjamin Becker. 2021. A distributed fMRI-based signature for the subjective experience of fear. Nature Communications 12 (1): 6643. https://doi.org/10.1038/s41467-021-26977-3. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. 1 3