Crime Victimization, Trust and Performance Evaluation of Donald Trump: Exploring Gender and Race Determinants

Abstract Two of Donald Trump administration’s most salient legacies are the high level of crime victimization and low popularity ratings among groups likely to be targeted by crime, such as females and race and ethnic minorities. To explore crime victimization and public opinion towards the president, the article tests whether gender and race determinants among crime victims were associated with trust in the president and job performance approval. Results indicated that females and female victims of crime were less likely to reward presidential performance and trust Trump. Blacks and Latinos were less likely to trust Trump and less likely to reward his job performance; however, there was no effect when moderating by whether they reported being victims of crime. Victims of crime in unsafe neighbourhoods were less likely to trust and reward presidential performance. More surprisingly, being a crime victim was a significant and positive predictor of trusting Trump and rewarding his job performance. The results tell us about the need to continue studying the public’s wishes for the authorities to respond to the problem of victimization, oftentimes rewarding populist approaches to punitive justice, incarceration and overall tougher criminal policies.


INTRODUCTION
During Donald Trump's four years in the presidency, the United States ranked consistently behind other industrialized democracies in overall peace measures including high perception of criminality, crime victimization, gender and racial discrimination, police militarization, and mass incarceration together with poor prison conditions (Gallup 2020a; Institute for Economics & Peace 2020; Lauritsen and Lentz 2019). Trump's presidency alone did not lead to these dismal and declining conditions; however, violent crime rates, such as serious assaults with gender and race implications, remained unvaried despite Trump's tough-on-crime rhetoric (Lopez 2020b). The country ranked third in serious assaults in the Americas in 2018, behind Colombia and Brazil. Homicides stalled at five per 100,000 population, and other indicators such as robbery, kidnapping, and the rates of persons arrested, prosecuted and convicted did not see a convincing decrease either (United Nations Office on Drugs and Crime; UNODC 2020). On average, between 2010 and 2018, 15% of the population were reported to be a victim of crimes, defined as robbery, burglary, assault, fraud, blackmail, extortion or violent threats (AmericasBarometer 2019a). Trump made several promises to the public which he sought to enforce while in office. He signed the bipartisan First Step Act to reduce recidivism, promoted networked governance between state and local law enforcement, supported nationwide police forces and mandated strategies to tackle gangand drug-related violence (White House 2020a). Yet, Trump belittled police brutality victims and rejected that some communities suffered more than others, called racial justice and equality movements such as Black Lives Matter "toxic propaganda", and claimed that immigration led to higher crime (Niedzwiadek 2020). In response, thousands of citizens joined nationwide protests calling for law enforcement reform and social justice (Misra 2019).
Trump's approval averaged from January 2017 to December 2020 at 41% with a base of predominantly White males and wealthy supporters (Gallup 2020b). Despite his track record of mockery and openly sexist and racist attacks to detractors, political opponents and lay citizens, some of Trump's most prominent criminal policy efforts were publicized as race and gender inclusive (Lopez 2020a;Shear and Sullivan 2018). The general public saw him as a stronger and more decisive leader than Joe Biden, although he got less support for being honest, having good judgement, showing a greater capacity to manage the government, and keeping his promises (Jones 2020). Trump threatened his female opponent, Hillary Clinton, in 2016 with imprisonment after the election, something his fanatics celebrated as a war chant: "lock her up!" Yet, the 2020 election revealed that racial minorities and women shifted their support toward Trump (Al-Gharbi 2020). Trump challenged the status quo condition, favouring conservative communities willing to vote for a radical candidate (Bang and Marsh 2018;Berman 2021;Fenger 2018). Why will minority groups support Donald Trump, and what was the relationship between being a crime victim in the Trump era? The article aims to theorize on Trump's popularity and examine victimization trends to understand how experiences with crime shape public sentiments. This line of arguing will lay out specific challenges in how society, especially gender and race minorities, put their confidence in the president and evaluate his job performance. The article poses a series of questions: What is the relationship between race and gender and being a victim of crime? Do victims of crime reward or punish presidential performance? What is the difference between males and females and between White and other race and ethnic minority victims of crime in their confidence in the president? Is neighbourhood insecurity associated with trust and approval?
To explore patterns of presidential approval among crime victims, the article is organized as follows. The first section gives Donald Trump's views on crime, the criminal justice system and crime victims. Reports on crime victimization in the years leading to 2020 are highlighted to indicate the potential gender and racial gaps that the article wishes to explore. The second section links the scholarship on crime victimization and presidential trust among public opinion. The third section explains the data selection from the 2019 AmericasBarometer and presents a set of testable assumptions. The fourth section unfolds the regression results and evaluates the likelihood of the hypotheses. The final section explains the evidence-based legacies of the Trump administration and offers practical elements for future research on public opinion, presidential expectations and crime victimization.

CRIME, VICTIMIZATION AND PRESIDENTIAL RHETORIC
During his campaign, Trump used gendered and racial discourses, arguing that rampant and uncontrolled criminality had stolen American values from the public (Donohue 2017). Once in office, he would meet with grassroots groups such as Black Voices for Trump and Women for Trump, claiming that the "worst crime conditions" that Black communities suffered resulted from previous democratic governments' miscarried policies. Hosting African American leaders at the White House, for example, Trump went on to say: We're here with some of the black leaders of our country andpeople that are highly respected and people that have done a fantastic job and, for the most part, have been working on this whole situation with me right from the beginning. And we've done a lot. We've done Opportunity Zones. We've done criminal justice reform. We've done things that people didn't even think possible. Criminal justice reformwe've let a lot of great people out of jail. (White House 2020c) Trump's crime narrative argued that the United States was experiencing "the lowest crime numbers in our country's recorded history" (Trump 2020). Yet, such a goal was far from accomplished.
According to the National Crime Victimization Survey (NCVS), violent-crime victims aged 12 years or older rose from 2.7 million in 2015 to 3.3 million in 2018, increasing by 604,000 victims. The rate of violent victimization excluding simple assault increased from 7.7 victimizations per 1000 females aged 12 years or older in 2017 to 9.6 per 1000 in 2018. The NCVS also showed that the offender was of the same race or ethnicity as the victim in 70% of violent incidents involving Black victims, 62% of those involving White victims, 45% of those involving Hispanic victims, and 24% of those involving Asian American victims (Morgan and Oudekerk 2019).
Following George Floyd's death, an African American detained by a White police officer who knelt on his neck for eight minutes and 46 seconds, Trump superficially commented on police reforms cornered by public demands for change. Flanked by White police officers, Trump avoided commenting on #MeToo and Black Lives Matter and rejected backtracking on some of his most inflammatory policies, including a negative to defund or dismantle police forces. "Without police, there is chaos," Trump argued as he chastised the looting and vandalism following Floyd's death (White House 2020b). More alarmingly, other episodes showed Trump's unapologetic attitude to unlawful behaviour from his supporters. He stood by Kyle Rittenhouse, a White 17-year-old militia vigilante who shot dead two Black Lives Matter protesters, saying he acted in self-defence. Later in his presidency, he showed no remorse after inciting a mob of White supporters to storm the Capitol as lawmakers met to certify the results of the 2020 presidential election. The riots, which left five people dead, became to some observers an example of White oligarchic violence in the country (Sachs 2021).
Before the COVID-19 pandemic shut down most of the country, civil society initiatives, including Black Lives Matter, rallied physically and online against the legacies of White hegemonic violence on non-White people, including police brutality and gender and race discrimination. States led initiatives to reform policing and criminal and social policy in Minneapolis, Atlanta, San Francisco, Los Angeles, New York and Chicago. A stoic Trump condemned the "liberal states", claiming that White people were also victims of police brutality. Against the president's claims, statistics produced by federal authorities revealed that African Americans are more likely to get fatally shot, get arrested at a higher rate for drug abuse, and get imprisoned more than other ethnic groups (Carson 2020; Federal Bureau of Investigation 2019; Tate, Jenkins, and Rich 2020). Race commentators depicted George Floyd's death, as well as the constant abuse of racial minorities, as the outcome of the regular criminalization of Black life in America (Pitner 2020).
Recent data show the ethnicity and race factor to be deeply central to the country's current state of politics and justice (Cobbina, Owusu-Bempah, and Bender 2016;Isom Scott 2017;Parks and Nowacki 2021;Peck 2016). According to a study by Baker, Perry, and Whitehead (2020), some of the strongest predictors of voting for Donald Trump in the 2016 presidential election were attachment to Christian nationalism and antipathy toward Muslims and immigrants. Unsurprisingly, these factors remained significant predictors of intention to vote for Trump in 2020, although identifying as African American had a negative but strong correlation.

Social Issues and the Public
Trump and the public usually clashed on their views towards many social issues, including crime. Data from the 2017 round of the World Values Survey indicate that 67% of people surveyed in the United States disagreed or thought it was hard to say that immigration led to more crime (Haerpfer et al. 2020). Most citizens had a favourable view of the #MeToo movement by September 2019 (52% versus 35% of unfavourable views). However, those identified with the Republican party tended to believe in men's denials over women's allegations in cases of sexual harassment and assault (48% believed women versus 91% of those identified with the Democrats). Most individuals disagreed with statements to put some fault on sexual assault victims. Republicans were more likely than others to agree with the sentiment that many women who are victims of sexual assault would not have been victims if they had been more careful (PerryUndem 2019).
Trump, who denied accusations by women of sexual misconduct, consistently reached out to gender minorities, for example, in his appointment in September 2020 of another female Supreme Court justice to replace the liberal Ruth Bader Ginsburg. Amy Coney Barrett's choice caused a stir among woman's rights defenders considering her social conservative background. Trump pushed controversial social policies thanks to the Republicans controlling Congress, and the Supreme Court with a conservative majority. According to Bernhard and O'Neill (2019:321), right-wing politicians opted to turn a blind eye to Trump's attacks and abuses against the independence of the rule of law.
Despite growing criticism by his opponents, Trump claimed that his justice and community protection policies were an "historic action to deliver a future of safety and security for Americans of every race, religion, color, and creed" (White House 2020b). Studies show a racially divided and partisan country. According to DeGue, Fowler, and Calkins (2016) studying the police's lethal force between 2009 and 2012, more White Americans have been killed by the police, year on year. However, the number of Black African Americans killed is 2.8 times higher than that of White people (Peters 2020).
Following Floyd's death, overall confidence in the police declined to decade lows as several deaths of Black Americans at the hands of armed officers ensued, including Breonna Taylor and Rayshard Brooks. The police's confidence ratings nevertheless rose among Republicans seven percentage points to 82% from 2019 to 2020, and dropped six points among Democrats to 28% (Brenan 2020a). What is the relationship between race and gender and being a victim of crime with trusting the president and rewarding job performance in the Trump era? What is the difference between males and females and between Whites and other race and ethnic minority victims of crime in their confidence in the president? In the next section, the article seeks to address these questions.

THEORETICAL CLAIMS: GENDER, RACE AND TRUST
Public opinion's trust in the president and rating job performance can depend on the effectiveness and responsiveness of implementing policy. Public opinion evaluates the personal characteristics of who gets appointed and how they deal with politicization and centralization of power (Resh 2015). Trust and performance evaluation can improve as presidents keep their promises while in office. Having a good economy, for instance, has proven to help governments in providing policies without requiring sacrifice from the public (Hetherington 2004). As citizens feel relatively well and public approval does not decrease, presidents propose policies that benefit some groups in society. On the other hand, the public is more likely to relate negative experiences in their lives that demand governmental performance with low levels of political trust. It is also associated with pessimistic predictions that government will solve their situation in the short term (Nicholls and Picou 2013). Most of the time, public opinion finds a balance that partly explains these two approaches.
Current challenges to democracy worldwide include low levels of electoral participation, slim trust in political parties, decaying accountability and transparency, and, most concerning to this article, declining trust in politicians and governance (Stoker and Evans 2019;Wike and Fetterolf 2018). Aspects of the current "democratic malaise" (Escobar and Elstub 2019) have eroded social capital in institutions that deal with social issues, especially the criminal justice system, making those affected by it lose confidence in the authorities.
Populist leaders can affect policy-making processes, and public opinion can influence their decision-making on critical social issues (Biard 2019). In Trump's case, research has shown how his unusual style put him to face many challenges at the core of government and in crucial policy areas, including criminal justice (Pfiffner 2017). Public opinion towards government authorities on crime and victimization issues remained low, as did the overall decline in trust in government. Potential causes for people's disaffection included partisanship and polarization, revealing different levels of trust in authorities and of confidence in the government (Citrin and Stoker 2018). The economy's recovery and a sharp decrease in unemployment since the recession of 2009 seemed to have helped Trump sustain public support at the end of his mandate. Nevertheless, the long boom pushed both antistatist and pro-statist populist policy solutions (Dunleavy 2018), which ultimately affected growing inequities, among them crime victimization.

Crime Victims and Political Preferences
The role of gender, race and class in American political behaviour has picked up interest among scholars trying to explain Trump's effect on social determinants as predictors of political preferences and participation (Baker et al. 2020;Harris and Rivera-Burgos 2021;Setzler and Yanus 2018). Theoretically, the public sees the personal integrity of those who represent the highest office in government as a prerequisite for effective democracy and leadership. However, the scandals, passive approval, omission of responsibility, and dishonesty performed by government authorities have tarnished trust in politics (Hosking 2014), resulting in widespread distrust of the head of state.
Because prominent features of Donald Trump's campaign and presidency included gendered and racially inflammatory rhetoric and fear over crime to incite the electorate (Newman, Shah, and Collingwood 2018), women and non-White victims of crime are expected to be less likely to trust and approve Trump. Women and non-White racial groups might not trust the president if they do not encounter safety nets in adversity such as crime.
This article builds on the argument that right-wing populism utilizes crime to make itself more appealing and acceptable to the public (Hosking 2019). In a heated televised debate in September 2020, for example, candidates Trump and Biden clashed over each other's track record of fighting crime, proposing police reforms, and condemning issues such as White supremacists and militias, questions which seemed to irritate Trump. The president continuously used his re-election effort to claim that his opponent was too soft and weak on crime.
A Gallup public opinion survey showed that public perception of crime was by then at its highest. Around 78% of the public thought there was more crime than before, against 64% in 2019 and 60% in 2018. Of the public, 63% thought that addressing social problems (i.e. putting more money to battle social and economic issues such as drug addiction, homelessness and mental health) felt closer to their views on fighting the crime issue, against a 34% who thought giving more money to strengthen law enforcement was the solution. Another 49% of the public said they were somewhat to very dissatisfied with the national policies to reduce crime, against 47% who said to be very to somewhat satisfied (Gallup 2020a).
Another theoretical claim is that public opinion can overlook the candidates' poor record of confronting crime or even accusations and convictions of committing crimes. Paradoxically, in the 2020 election, the states ranked with the highest rates of violent crime mainly voted in favour of the Republican president: Alaska (Trump was declared the winner); New Mexico (Biden); Tennessee (Trump); and Arkansas (Trump). Meanwhile, the states with the lowest rates in violent crime all went for the Democratic candidate: Maine (Biden); Vermont (Biden); New Hampshire (Biden); and Virginia (Biden) (Stebbins 2020).
Although the 2018 mid-term election was celebrated for reaching record numbers of women and ethnic minorities elected, voters also re-elected two Republican congress members in New York and California facing federal indictments for alleged insider trading and campaign corruption, a Democratic senator on trial on federal bribery and fraud charges, a Texan attorney general facing criminal charges for securities fraud, and a congressman in Montana convicted for assaulting a journalist (Time 2018). Maybe the most relevant claim was proposed by Hart (2020), arguing that Trump won people's feelings by swinging public opinion beyond issues of partisanship, policy or economic factors (Warren-Gordon and Rhineberger 2021). Similarly, research by Schaffner (2020) claimed that exposure to Trump's bigoted rhetoric stimulated prejudice, making people express more bias to race and gender gaps. "Promises to 'Make America Great Again' refer to an American tradition that subjugates African Americans to a second-class citizenry and that will not disappear with an incendiary president," claimed Gause (2018), critically reflecting on the relationship between Trump and African American voters, the electoral participation of woman, transgender and non-White candidates since 2016.
A recent finding suggests the existence of a Trump effect, with negative perceptions among minorities relating to how his presidency showed that the United States is a racist society, straining police-minority relationships and normalizing an environment in which racial minorities are likely to be arrested or subjected to police violence (McManus et al. 2019). For some scholars, Trump's presidency became harmful to vulnerable communities, either on a low income or oppressed by their race or gender, on specific policies in the areas of criminal justice, voting rights and environmental discrimination (Clayton, Moore, and Jones-Eversley 2019).
Regarding gender issues on the public agenda, the evidence presented by Kulig, Cullen, and Haner (2019) reveals that 51% of US respondents in a commissioned YouGov survey believed that "the women told the truth when they reported that Mr. Trump sexually harassed or assaulted them" versus 23.1% who sided with the president. The questions remain why the citizenry diverges on moral narratives towards Trump being either an honest president or a predator. Thus, a next idea would be that patriarchal and misogynistic attitudes and language toward women became a normal feeling under Trump. Scholars argue that Trump opened a path of devaluing, demonizing and objectifying minorities that triggered hostile rhetoric favouring White, heteronormative elites (Rothe and Collins 2019).
Divided citizens' confidence in the authorities leading the crime issue shed light on the partisan state of public affairs. Trust in the US criminal justice system, for example, varies by party identification and racial background. A 58% majority of Republicans and Republican-leaning independents say the criminal justice system is not tough enough. However, only 25% of Democrats and Democratic-leaning independents shared this view. More White Americans than non-White Americans said the justice system was not tough enough on crime (45% versus 31%, respectively) (Brenan 2020b).
Research examining economic stress and differences in victimization rates indicates that cultural context influences whether people below or above the median income are more frequently crime victims (Neapolitan 2003). Scholars have shown that in some US cities, for example, the national homicide increase was concentrated in communities that previously had the highest violence rates (Lauritsen and Lentz 2019). Elsewhere, the study of the relationship between crime victimization and neighbouring behaviours has indicated that the victimization's negative impact on place attachment is attenuated through frequent social interactions with neighbours (Zahnow and Tsai 2019). For example, daily and routine activities are significant predictors of females' risks; meanwhile, neighbourhood conditions seem to be better indicators of males' risks (Like-Haislip and Miofsky 2011). Residential security decreased the chances of violent victimization for White and Latino males; however, it increased the likelihood of Black males' victimization (Like-Haislip and Miofsky 2011). Table 1 shows that males, Whites, those over 65 years old, and those earning over $5000 per month had the largest approval rates for Donald Trump. Support for Trump did not vary in great depth across his time in office despite his poor record on confronting crime, the low public confidence in the criminal justice system and the police, or even being impeached on charges of abuse of power and obstruction of Congress. Pollsters registered 45% of approval by January 2017 at the beginning of this presidency and 39% of support a month before leaving office in December 2020. What is less surprising but worth noting is the 87-percentage-point gap in the president's approval rating between Republicans (91%) and Democrats (4%) recorded in July 2020 (Gallup 2020b).
In sum, demographic factors and personal beliefs about Trump's role in American politics remain a predictor of bias among women and men, Whites and non-Whites across a range of issues. While most Trump voters in 2016 were male, more than 40% of women who voted also supported him. Setzler and Yanus (2018) found that party affiliation, sexism and racial resentment were significant predictors and had a greater influence on Trump voters of both genders.

HYPOTHESES AND DATA
This section explains the data selection and presents a set of testable hypotheses. The article uses trust and performance rates as dependent variables and asks what determinants improve or lower confidence and approval towards the president. It claims that when gender and racial minorities are respected, they should trust and reward the president; when leaders do things poorly, they do not. As other scholars have argued, individual-level variables (i.e. age, gender, education, income, etc.) can be better predictors when studying crime victimization outcomes than macro-level variables related to political and economic conditions (Uludag et al. 2009).
The article uses race and ethnicity as a sole categorical variable due to the data extracted for the analysis. However, they should be understood on their own as personal identifiers (race) and as learned identities (ethnicity).
Following the literature reviewed above, the article tests the following set of hypotheses: H 1 gender: Female victims of crime are less likely to trust the president and less likely to reward presidential performance.
H 2 race and ethnicity: Non-White victims of crime are less likely to trust the president and less likely to reward presidential performance. H 3 neighbourhood security: Victims of crime in unsafe neighbourhoods are less likely to trust the president and less likely to reward presidential performance.
H 4 household income: Victims of crime earning below the mean of household income are less likely to trust the president and less likely to reward presidential performance.
The hypotheses above aim to compare the relationships between the key independent variables and presidential approval and trust within the samples of crime and non-crime victims. The article uses the version of the AmericasBarometer conducted in July 2019 in the United States by YouGov. The AmericasBarometer is useful and particularly relevant as it includes demographic, socio-political, security and victimization questions in one dataset. To measure victimization, the article included the question: "Have you been a victim of any type of crime in the past 12 months? That is, have you been a victim of robbery, burglary, assault, fraud, blackmail, extortion, violent threats, or any other type of crime in the past 12 months?" To measure the dependent variables about confidence in the president, the question used was: "To what extent do you trust the President?". This was a continuous measure taken from the original scale coded 1 (strongly disagree) to 7 (strongly agree). To measure the dependent variables about confidence in the president, the question used was: "How would you rate the job performance of President Donald Trump?" The original variable was reverse coded in a five-point Likert scale with values going from 1 (very bad) to 7 (very good). To test hypotheses 1 and 2, a survey question asking for self-identified gender was used (recoded females = 1), and race (recoded into three separate dummy variables with base categories being Blacks, Latinos, and other race). To test hypothesis 3 (perception of neighbourhood security), the question was used: "Speaking of the neighborhood where you live and thinking of the possibility of being assaulted or robbed, do you feel very safe, somewhat safe, somewhat unsafe or very unsafe?" This variable was recoded (unsafe = 1, and safe = 0) from the original values asking participants whether they felt safe, somewhat safe, somewhat unsafe, or very unsafe. Testing hypothesis 4 about household income required using a question in which participants were asked for their household's total monthly income, including remittances from abroad and the income of all the working adults and children. Household income was initially measured using a 17-point scale ranging from no income to more than $17,000. A dummy variable was created with two groups including those with an income of less than or equal to the mean range (i.e. between $2201 and $2700; this group was coded 0), and those with an income above that range (coded 1). Table 2 presents the variables' descriptive summary statistics.

RESULTS
The sample (N = 1500) included voting-age adults from the general population (aged 19 to 91 years). The sample was drawn using sample matching to select representative samples from non-randomly selected pools of respondents. To produce the final dataset, YouGov interviewed a larger pool of opt-in panellists and then From the surveyed sample, 232 individuals (15.47%) reported having been victims of crime in the 12 months before being interviewed. Of the crime victims, 118 were females (50.8%) and 114 were males (49.1%). Victims of crime grouped by race showed that the majority were White (56.03%) followed by Hispanic or Latino (18.53%), Black or African American (15.52%), Asian American (3.88%), mixed race (2.16%), Native American (2.16%), other (.86%) and Middle Eastern (.43%). Pearson's χ 2 test for the cross-tabulation of race and ethnicity and being a crime victim has 7 degrees of freedom and is 38.85. The observed differences are statistically significant at p < .001.
The ordinary least squares (OLS) regression results are shown in Table 3. Consistent with previous studies, women, Blacks, Latinos and those feeling the neighbourhood was unsafe were less trustful of the president (Model 1), whereas those who were crime victims were more trustful. Identifying as being Black and female had the largest effects. Household income and being from another race were not statistically significant. For this model, being a crime victim resulted to be a significant and positive predictor of trust in the president. The effects of gender and race on presidential trust was not moderated by perceiving an unsafe neighbourhood (Model 2). The main effects of gender, race and neighbourhood safety were negative and significant, and crime victimization was positive and statistically significant; however, the interactions drew mixed results, with only females thinking that their neighbourhood was unsafe, having a negative and statistically significant effect. The effect on the trust outcome of living in an unsafe neighbourhood was moderated by victimization, with a negative and statistically significant coefficient. Figure 1(a) presents these results by predicting average scores on trust in the president scale varying by whether someone was a crime victim by gender identification, holding all the other variables at their means. Being a male increased scores regardless of crime victimization, but males that reported being a crime victim scored .77 points higher in the scale than non-victims. For females, the predicted scores on trust were lower than males, but, again, crime victims scored higher in the trust scale than non-victims, although by a shorter difference of .36 points in the scale. Table 3 also shows the regression results on attitudes towards Donald Trump's job performance (Model 3). Women, Black and Latino people, and believing the neighbourhood was unsafe were negatively and statistically significant predictors of job performance scores. Like Model 1 (without interactions), having reported   Source: Author's construction with data from AmericasBarometer (2019a). Note: Standard errors are in parentheses. †p < .08, *p < .05, **p < .01, ***p < .001.
being a victim of crime became a positive and significant predictor of presidential job performance. In contrast to Model 1, race factors showed the strongest effects (being Black and Latino), this time above the case of being a female. The effects of gender and race were also moderated by feelings of unsafety in the neighbourhood and crime victimization (Model 4). Like the findings in Model 2, the results of Model 4 (with interactions) revealed that females that were victims of crime gave lower scores to Trump's performance. This relationship was the only reported with a statistically significant effect. The effect on attitudes towards Trump's job performance of living in an unsafe neighbourhood was moderated by victimization, with a negative and statistically significant coefficient. Figure 1(b) conveys the predicted average scores of presidential job performance scores and whether a person has been a victim of crime. Performance scores are higher for crime victims regardless of gender. In the case of males, the difference was in favour of having been a victim, scoring .75 points higher in the scale. Meanwhile, for females, the difference in groups was much lower by only .25 points.
In sum, females, and female victims of crime were less likely to reward presidential performance and trust in the president (H 1 gender, accepted). Non-Whites are less likely to trust the president and less likely to reward presidential performance, however, there is no effect when moderating by whether they reported being victims of crime (H 2 race and ethnicity, rejected). Victims of crime in perceived unsafe neighbourhoods are less likely to trust and reward presidential performance (H 3 neighbourhood security, accepted). Being a victim of crime earning below the mean of household income has a null effect in trusting the president or approving presidential performance compared to earnings above the household income mean (H 4 household income, rejected).

CONCLUSION
Complementing previous studies, the article concludes that crime victims' gender, race and ethnicity are important moderators of public attitudes (Cramer 2020;Harris and Rivera-Burgos 2021;Uggen and Larson 2017;Ziv, Graham, and Cao 2019). The article has shown that the experience of being a female victim of crime increases the effect of not trusting and not approving the president's job performance. When accounting for belonging to a racial and ethnic identity other than White, such as being Black, Hispanic, or Latino, or other race, these significant effects of these determinants moderated by crime victimization and perception of unsafe neighbourhoods disappear. Still, the moderating effects of being a victim of crime and living in an unsafe location are strong predictors of mistrust and believing the president had a bad job performance. More surprisingly, being a victim of crime was a significant and positive predictor of trust in Trump and rewarding his job performance. These results speak to the literature exploring the varied role in the public's wishes for the authorities to respond to the problem of victimization, many times rewarding populist approaches to punitive justice, incarceration and overall tougher criminal policies (Kort-Butler and Hartshorn 2011; Lawton and Clark 2015; Thompson and Bobo 2011).
Some scholars suggest that public expectations in the presidency tend to idealize politics and culture around images of personal traits and desirable qualities (i.e. honesty, sound judgement, competence, confidence, etc.). Performance-based expectations represent public beliefs that the president's office is equipped to maintain peace, security, prosperity, economic growth and social justice (Gerber et al. 2011;Simon 2010:142). Linked to this body of literature, the article has shown that victims of crime in the United States have both image-based and performance-based determinants that can condition support for presidential trust and job performance. As Smith (2019) argued, Trump partisans were particularly fond of an authoritarian leader who would crush rivals at home and abroad and mainly use his powers to deter feminists, liberals, immigrants and minorities. Grasping attitudes among crime victims become essential to understand presidential support.
It could also be argued that social attitudes towards politicians and trusting and rewarding the president's figure can co-exist with the fact of being victimized by crime. Gender and race factors, however, help explain the complexity of society, culture and political life in the United States. The findings here suggest that the Trump administration's misogynistic attitude and language toward women (Rothe and Collins 2019) have consequences on his approval and more so across female crime victims across the country. The results of this article also confirm previous work, exploring the damaging effects of the Trump administration on the increased racial divisions and the erosion of faith in institutions across American society (Conley 2020;McClain 2021;Parker and Towler 2019).
Finally, the level of the jurisdiction where most crime issues are handled needs further research to compare federal efforts and local governance of crime at the state and county levels. During his single term in office, Trump appointed three justices to the US Supreme Court, 54 United States Circuit Court of Appeals judges and 174 federal District Court judges, often frustrating many of his administration officials' efforts to enact and implement judicial policy (Nemacheck 2021). Although some states (i.e. Florida) adopted constitutional amendments protecting crime victims' rights (Cassell and Garvin 2020), it is up to the head of the state to consider crime victims' trends, especially of the most affected population groups that transcend federal polity. Crime victims will distrust and punish nationally elected officials if they ignore their daily struggles. This requires a greater conversation on and beyond what the legacy of Donald Trump will be for American society.