Asian Australians’ Experiences of Online Racism during the COVID-19 Pandemic

: Between 13 November 2020 and 11 February 2021, an online national survey of 2003 Asian Australians was conducted to measure the type and frequency of self-identiﬁed Asian Australians’ experiences of racism during the COVID-19 pandemic. The survey also aimed to gauge the relationships between racist experiences and targets’ mental health, wellbeing and sense of belonging. In this paper, we report ﬁndings on the type and frequency of online racist experiences and their associations with mental health, wellbeing and belonging. The survey found that 40 per cent of participants experienced racism during the COVID-19 pandemic. Within that group, 66 per cent experienced racism online. The demographic pattern of those most likely to experience online racism were younger age groups, males, those born in Australia, English speakers at home, non-Christians, and migrants who have been in Australia less than 20 years. Analysis also found a strong correlation between Asian Australians’ experiences of online racism and poor mental health, wellbeing and belonging. The relationship between experiencing racism, non-belonging and morbidity were more pronounced for those who experienced online racism compared to those who experienced racism in other ofﬂine contexts. This points to the corrosive nature of online racism on social cohesion, health and belonging.


Introduction
In January and February 2020, one in four people who reported racial discrimination to the Australian Human Rights Commission linked the discrimination to the COVID-19 pandemic. In February, the Commission also recorded the highest monthly number of racial discrimination complaints in that financial year (Fang et al. 2020). On 8 April, the Race Discrimination Commissioner, Chin Tan, announced that the 'persistent reports of racist abuse and discrimination' since the beginning of the COVID-19 outbreak is 'an alarming trend that deeply affects individuals and undermines the community's ability to respond effectively to the pandemic' (Tan 2020a). Mr. Tan advocated for a national response to racism arising from the pandemic, including the 'comprehensive collection and evaluation of data' (Tan 2020b) and backed an open letter from prominent Chinese Australians calling for national unity and calling out racism, particularly against people of Chinese background (see Yat-Sen Li et al. 2020). In the open letter, signatories declared that they are 'deeply concerned that the recent rise in anti-Chinese sentiment is driving a marked escalation in racial abuse towards Asian Australians. This poses a serious threat to our national unity'. The letter referred to 'footage of the vilification of Asian Australians that has circulated globally across social media' and declared that the 'disturbing trend' of Australians 'being targeted because of their Asian heritage or appearance' cannot 'continue unchallenged'.
As the reference to global social media in the open letter suggests, anti-Asian sentiment linked to the COVID-19 pandemic was not limited to the Australian context. Internationally, Important information has been obtained from these studies regarding the nature and extent of racism during the pandemic. However, there are limitations to the existing data associated with sample sizes, sampling frames and survey designs. More specifically, these projects have not been able to provide empirical evidence that the prevalence of anti-Asian racism has indeed escalated during the COVID-19 pandemic as per anecdotal and media reports. This is similarly the case with many international studies. An exception is the social media (Twitter) study by Nguyen et al. (2020) which indicated a marked increase in anti-Asian sentiment and a decline in positive Asian sentiment in 3.4 million tweets following the emergence of COVID-19 in the U.S. The procurement of rigorous empirical evidence of Asian Australians' experiences of racism during the COVID-19 pandemic (the nature, type and frequency), how this has changed over time (pre-and since the pandemic), and the potential impacts of such experiences, is therefore in need of continued and urgent attention.
The study presented in this paper sought robust empirical evidence to gauge the extent and form of racism occurring during the pandemic (and how this compares to the pre-COVID-19 context), and uncovers particularly vulnerable groups within Asian Australian communities. These overarching aims were guided by the investigation of the following: (1) the type and frequency of experiences of racism and how they may vary across the Asian Australian subgroups; (2) the potential relationships between these racist experiences and targets' mental health, wellbeing, and sense of belonging; and (3) how targets and witnesses of racism respond to such racist incidents/experiences (e.g., official reporting, seeking information, peer or family support, and/or counselling). For the purposes of this study, 'type of racism' referred to interpersonal experiences of racism such as being treated less respectfully, not being trusted, being called names, ridiculed or insulted, being physically assaulted and being denied service (as per Blair et al. 2017;Markus and Dharmalingam 2007). 'Type of racism' also included an investigation into the settings in which racist experiences occurred-such as in shops, on public transport, in dealings with the police or finding housing etc. (as per Blair et al. 2017;Markus and Dharmalingam 2007). Mental health and wellbeing were defined in accordance with the Depression, Anxiety and Stress Scale (Lovibond and Lovibond 1995), the Brief Resilience Scale (Smith et al. 2008), Index of Race-Related Stress (Utsey and Ponterotto 1996) adapted in the LEAD Experiences of Racism Survey (Ferdinand et al. 2012), Experiences of Discrimination Scale (Krieger et al. 2005) and Williams' unpublished racism-related vigilance scale reproduced in Clark et al. (2004) and were also adapted in the LEAD Experiences of Racism Survey (Ferdinand et al. 2012). Sense of belonging included perceptions of interpersonal connections such as being included, having close bonds with family and friends, as well as perceptions of feeling 'Australian' and the importance of being accepted into Australian society and continuing cultural practises. These 'sense of belonging' items were drawn from the General Belongingness Scale-acceptance/inclusion subscale (Malone et al. 2012) and the Challenging Racism National Survey (Blair et al. 2017).
The research aims and objectives were achieved via a national survey of Asian Australian adults. The methodology is detailed below. This paper focuses on the study findings relating to Asian Australians' experiences of racism in the online setting (rather than across other settings such as 'on the street', 'work', or 'public transport) during the first year of the COVID-19 pandemic (up until February 2021 1 ) and the relationship with Asian Australians' mental health, wellbeing and sense of belonging.

Racism in Online Settings
Online communication plays an integral role in day-to-day interactions. In their representative survey of over 3500 Australians, the Australian Communications and Media Authority (ACMA 2021a) found that 99% of participants used the internet in the 6 months to June 2021. This was an increase from 90% in 2019 and was likely accelerated by the COVID-19 pandemic where usage of online communication and services increased notably as remote working, study and provision of 'everyday' services such as shopping, education, health services, and banking shifted increasingly online (ACMA 2021a). Most Australian adults (81%) accessed the internet multiple times a day (ACMA 2021a) and, more specifically, 96% of participants used a communication or social media website or app for personal purposes (ACMA 2021b). While younger Australians were the greatest users of communication or social media websites/apps, all age groups (18+ years of age) were active users, with 76% of users actively engaging with content such as posting, sharing, commenting or 'liking' a post (ACMA 2021b).
Online settings such as communication and social media websites/apps have some level of anonymity, fleeting encounters and disembodiment that can manifest in disinhibition (Suler 2004) and less accountability (Hughey and Daniels 2013). Such features may provide a powerful means for racist attitudes and vilification to be enacted and propagated. In fact, Tynes (2007) suggested that online racism is experienced at the same or at higher levels to experiences of racism in the offline environment. More specific to the Australian context, the Australian Human Rights Commission has indicated that close to one in ten racism complaints prior to the COVID-19 pandemic were in relation to online racism (Jakubowicz et al. 2017, p. 67). Community organisations have also reported significant levels of online racism in regards to Islamophobia (Iner 2019), anti-Semitism (ECAJ 2020, and incidents reported by Asian Australians during the early months of COVID-19 (Chiu and Chuang 2020). While Volpe et al. (2021) revealed important findings regarding the prevalence and impacts of structural online racism (i.e., racism embedded in technological infrastructure that generate and reinforce inequities among racial groups); our focus is on interpersonal racism experienced online. Thus, in this paper we use a definition of online racism put forward by Bliuc et al., as "any form of communication via electronic or digital media by groups or individuals which seeks to denigrate or discriminate against individuals (by denying equal rights, freedom and opportunities) or groups because of their race or ethnicity" (Bliuc et al. 2018, p. 76).
The anonymity and transience facilitated in online settings can enable users to escape punishment or regulation from authorities (Piracha et al. 2019;Jakubowicz et al. 2017). Unregulated racism in online settings can be seen as a form of roll-back neoliberalism (Peck and Tickell 2002), where human interactions that were previously regulated become less regulated or unregulated due to anonymity and lack of redress. Online racism also has aspects of roll-out neoliberalism, which calls for individuals, rather than platforms or agencies, to carry the burden of monitoring and responding to racism (Hughey and Daniels 2013). For example, in the Australian context, a recent Victorian Parliamentary inquiry into anti-vilification protections dedicated a chapter to online vilification, and recommended collaborating with community organisations to improve data collection, increase online safety skills, and raise awareness about anti-vilification laws online (Parliament of Victoria 2021, pp. 221-54). Without adequate reporting and regulation that moves beyond individual responsibility, such burdens are most profoundly felt by the targets of racism online, typically from vulnerable groups such as Asian-Australians who are the focus of this paper.
While the effects of face-to-face racism on targets is well documented (Grigg and Manderson 2014;Paradies 2006;Paradies et al. 2015), there is less attention paid to the specific effects of experiencing online racism ). Online racism has been shown to intimidate, injure and degrade belonging and civic participation; it can also be a vehicle for racist social movements with the capacity to undermine social cohesion Ferdinand et al. 2015). While overt physical violence is less likely within online settings, there is evidence to support the reproduction of power hierarchies, the embodiment of structural violence on vulnerable populations (Carlson and Frazer 2018), and links between online communication and offline conflict (Bliuc et al. 2019). A systematic review of studies into cyber racism conducted by Bliuc et al. (2018) argues that more academic attention is needed to investigate the effects of online racism on targets.
In response to Bliuc et al.'s (2018) findings and in alignment with the recommendations proposed by the Victorian Parliamentary inquiry into anti-vilification protections (Parliament of Victoria 2021), this paper provides empirical evidence for the relationship between online racism and mental health, wellbeing, and a sense of belonging. The focus on the experiences of Asian Australians during the COVID-19 pandemic is particularly pertinent. During the pandemic, a large portion of people's lives moved online due to lockdown restrictions, though we acknowledge this was applied unevenly across socioeconomic and socio-cultural axes. There is also emerging research to support the varied ways in which racism (particularly against those of Asian ethnicity/background) was enacted through online platforms during the pandemic; for example, racist zoombombing (Lee 2021), racist sentiment and emotion in tweets (Dubey 2020) and the racist and discriminatory online responses to COVID-19 that disproportionately affected marginalised groups (Devakumar et al. 2020). The empirical evidence provided in this paper therefore provides important, timely, and urgent insights. We argue that without adequate data, the full impact of online racism-and how it undermines social cohesion and individual mental health and wellbeing-remains hidden.

Method
This project utilised an online national survey of Asian Australians between 13 November 2020 and 11 February 2021. The Australia-wide online survey comprised a sample of 2003 self-identified 'Asian Australians' 16 years of age or older currently residing in Australia. The sampling frame was provided to the online panel provider Dynata, who have a pool of approximately 300,000 Australian panelists. Dynata employs rigorous quality controls, recruits participants in a variety of ways to ensure a diverse sample, and randomly selects panel participants to complete the surveys. It is acknowledged that there may be various limitations to utilising online panels in survey research, such as the relationship between completion rates and data accuracy and the relationship between multiple panel memberships and data quality; however, these limitations are not yet well examined (Callegaro et al. 2014). One obvious limitation to this survey design is that participants must have access to the internet and must be registered panelists of Dynata.
Despite these recruitment limitations, the final sample for this project were diverse across all demographic variables. For example, of the total sample, 56.6% were female, 42.7% were male and a small number (0.4%) were non-binary/gender fluid. All age categories (16+) were represented in the survey sample, with quite a high representation of Asian Australians between 25 and 44 years of age (29.1% in both the 25-34 and 35-44 age brackets). As is typical in survey studies, there was a skew towards the more educated: 70.8% had a university degree or postgraduate qualification; 14.5% had other tertiary qualifications including a trade or TAFE (Technical and Further Education) qualification; 12.3% had year 12 or equivalent as the highest level of qualification; and 1.3% had no formal qualifications. The majority of participants were employed (65.2%). The unemployed were slightly over-represented at 7.5% compared to the January 2021 national rate of 6.4% (ABS 2021). Students (8.9%), those with caring/home duties (6.7%), and those that were retired (5.2%), were also represented. Approximately half of the survey sample earned above the average annual income of $65K (ABS 2019), with 24.3% of participants earning between $80,000 and $149,999 per annum. The sample included respondents from across all states and territories: Victoria (37.8%); New South Wales (36.7%); Queensland (10.6%); Western Australia (7.8%); South Australia (4.4%); Australian Capital Territory (1.7%); Tasmania (0.7%); and the Northern Territory (0.2%). Using participants' residential postcode information, both metropolitan 2 and non-metropolitan residential locations were represented (89.3% metropolitan; 10.7% non-metropolitan).
In terms of culture/ethnicity, citizenship and birthplace variations, the sample was similarly diverse. The majority (62.1%) of participants were born overseas. The most common overseas birthplaces were India (10%), Philippines (6.1%), Malaysia (6.0%), China (4.6%), Vietnam (4.3%) and Hong Kong (3.5%). Of the 62.1% of participants who were born overseas, the majority (29.1%) were residing in Australia for more than 20 years. This was followed by 25.7% who have been in Australia for 5-10 years, 22.2% for 11-20 years, 14.5% for 2-4 years, and 8.5% who have been residing in Australia for less than 2 years.
In line with the project's aims, the survey included closed-response and open-response questions that were used to obtain information relating to experiences of racism (type, frequency, changes over time), the actions of bystanders, responses to racist incidents and reporting, morbid effects of racism and impacts on wellbeing and resilience. These questions were largely drawn from the international standards in racism and psychology (mental health/wellbeing) research (including Markus and Dharmalingam 2007;Blair et al. 2017;Lovibond and Lovibond 1995;Smith et al. 2008;Utsey and Ponterotto 1996;Ferdinand et al. 2012;Krieger et al. 2005;Malone et al. 2012). Additional bespoke questions (particularly relating to the COVID-19 context) were developed where necessary. In this paper, we focus on the findings relating to experiences of online racism during the pandemic only, and in particular, their relationship to mental health, wellbeing and sense of belonging among Asian Australians.

Quantitative Analysis
Descriptive statistics (primarily frequencies) and bivariate analyses (e.g., cross-tabulations) were used to generate an in-depth quantitative understanding of experiences of racism among Asian Australians, and the extent to which these experiences varied across a number of demographic and social groups (e.g., age groups, gender, cultural/ethnic backgrounds, geography, Australian born vs. overseas born). Cross-tabulations were used to examine relationships between experiences of racism during the COVID-19 pandemic and mental health and wellbeing, mobility (e.g., avoiding situations), anticipation of racism and sense of belonging. A p-level of 0.05 was used to interpret the significance of associations between variables. By examining the descriptive statistics, the experiences of Asian Australians across various subgroups of the population were examined.

Qualitative Analysis
Open-response questions in the online survey invited further detail and explanation as to whether experiences of racism had increased in the COVID-19 context and also further information and detail regarding COVID-related incidents of racism experienced. Due to word-length limitations, presentation and analysis of the qualitative data are not presented in this paper.

Type and Frequency of Online Racism during the Pandemic
Of the full sample, 40 per cent (n: 798) of participants experienced racism in the COVID-19 context. Within that group (n: 798), 66.4 per cent (n: 530) experienced racism online (hardly ever/1-2 times-very often/12+ times). The online setting was also the most frequent setting in which racism occurred, with 15.1% (n: 120) of the sub-sample (those that experienced racism) having experienced online racism 'often/very often' (8+ times). This was at least 3.4 percentage points higher than the two other common contexts in which racism was experienced during the pandemic: in shops (11.7% often/very often) and on the street (10.8% often/very often). It must be noted, however, that the experience of online racism during the pandemic decreased from the pre-COVID-19 context. In the 12 months prior to the pandemic, 47.9% (n: 959) of participants experienced racism. Of this group, 18.8% (n: 180) experienced online racism 'often/very often' (see Table 1). While decreases in racism in 'offline' spaces could be explained by lockdown measures that limited social interactions, the reason for the decrease in experiences of online racism among participants during the pandemic is in need of further investigation. 1 Sample sizes reflect the fact that only participants who stated that they had experienced racism or racial discrimination were asked to provide details of context/settings. Moreover, note that the sample sizes of those who experienced racism before and during the pandemic differ. In the 12 months prior to the pandemic n: 959 participants experienced racism; during the pandemic n: 798 participants experienced racism. Question wording: 'In the 12 months prior to January 2020 (January 2019-January 2020), how often did you experience racism or racial discrimination in the following situations?' and 'Since January 2020, how often have you experienced racism or racial discrimination . . . " Response wording: Never (0 times), Hardly ever (1-2 times), Sometimes (3-7 times), Often (8-10 times), Very often (12+ times), Not Applicable (N/A). Question adapted from Blair et al. (2017).
The demographic pattern of those most likely to experience online racism were younger age groups, males, those born in Australia, English speakers at home, non-Christians, and migrants who have been in Australia less than 20 years (see Table 2). 1 Sample size is n: 374 as only migrant participants who experienced racism during the pandemic have been included. Due to the smaller sample size, responses 'hardly ever', 'sometimes', 'often', and 'very often' have been collapsed into one category 'hardly ever-very often'. Question wording: 'Since January 2020, how often have you experienced racism or racial discrimination . . . " Response wording: Never (0 times), Hardly ever (1-2 times), Sometimes (3-7 times), Often (8-10 times), Very often (12+ times), Not Applicable (N/A). * Statistically significant at p < 0.001. Question adapted from Blair et al. (2017).

The Relationship between Online Racism and Mental Health/Wellbeing
Respondents that experienced online racism during the pandemic (n: 530) reported worryingly poor mental health and wellbeing during the COVID-19 pandemic (see Table 3). In particular, these participants reported being stressed, anxious and depressed. For example, seven out of ten participants that experienced online racism had experienced breathing difficulty in the absence of physical exertion (anxiety), and 71.6% found it hard to wind down (sometimes-almost always) (stress). In addition, the vast majority (85.5%) of those who experienced online racism during the pandemic felt down-hearted and depressed as well as feeling they are not worth much as a person (85.5%) (depression). Among the respondents who experienced online racism during the pandemic, the proportion that found their resilience waning-or the ability to 'snap back' when something bad happens' difficult-increased during the pandemic. A little over one-third (36.4%) of this group found it hard to snap back when something bad happened prior to the pandemic. During the pandemic, finding it hard to 'snap back' increased to one-half (51.1%). The relationship between experiencing online racism and wellbeing is also apparent when viewing results for participants' anticipation of racism. Specifically, 71.5% of respondents that experienced online racism during the pandemic anticipate people saying or doing something racist when interacting with others (sometimes-very often), and 81.2% avoid specific situations (sometimes-very often) because of racism. In your daily life, how often do you try to avoid specific situations because of racism? 4.9 13.9 34.9 29.8 16.5 Questions adapted from Lovibond and Lovibond (1995); Smith et al. (2008); Utsey and Ponterotto (1996); Clark et al. (2004) and Ferdinand et al. (2012). Table 4 presents cross-tabulations of the frequency of experiencing online racism during COVID-19 with frequency of mental health and wellbeing indicators. The results show that Asian Australians who experience more frequent racism online report poorer mental health and wellbeing, as compared to those who experienced less frequent (or no) online racism. When viewing the mental health and wellbeing of participants that did not ('never') experience online racism during the pandemic (but did experience racism in other settings), the pattern is particularly stark. For example, 37.1% of those that 'never' experienced online racism had experienced breathing difficulty in the absence of physical exertion. This compares to 70% who experienced online racism (hardly ever-very often). Similarly, 70.5% of those who have 'never' experienced online racism felt downhearted and depressed (compared to 85.5% who experienced racism online), and 53.9% felt they were not worth much as a person (compared to 85.5% who experienced racism online). In terms of resilience, 36.3% of those who 'never' experienced online racism found it hard to snap back when something bad happened during the pandemic, compared to 51.1% of those who experienced online racism. This is an almost 15 percent difference between the two groups. Unfortunately, this relationship between frequent experiences of online racism and poorer outcomes is consistent across all the mental health and wellbeing indicators.   Lovibond and Lovibond (1995); Smith et al. (2008); Utsey and Ponterotto (1996); Clark et al. (2004) and Ferdinand et al. (2012).
Within the sub-sample of those who experienced online racism during the pandemic, males, middle/aged or older people, those born in Australia, English speakers at home, migrants who have been in Australia for over 10 years, and non-Christians were most likely to experience poorer mental health outcomes. In fact, there were significant demographic differences in 11 of the 14 mental health and wellbeing indicators. Due to space constraints, only selected significant demographic patterns will be highlighted here.
In relation to one of the survey's stress indicators, 25.3% of those aged 35-44 years (who had experienced online racism during the pandemic) 'almost always' found it hard to wind down, compared with 14.7% of those aged 45 years and over, 11.0% of 25-34 year-olds, and 7.4% of 16-24 year-olds. Differences in stress response was also seen in relation to the language spoken at home: 18.1% of those who spoke English and experienced online racism during the pandemic 'almost always' found it hard to wind down as compared to 8.0% of those who spoke a language other than English at home. Over one-third (36.4%) of migrants who have lived in Australia for 11-20 years and experienced online racism during the pandemic 'almost always' found it hard to wind down, compared to 7.7% who have lived in Australia for over 20 years, 8.3% who have lived in Australia for less than 2 years, 0.0% for those who have lived in Australia for 2-4 years, and 3.9% for those who have lived in Australia for 5-10 years. In regards to religion, 17.3% of those who were religious non-Christians 'almost always' found it hard to wind down, compared with 14.4% of Christians, and 10.5% of those who are non-religious.
In terms of anxiety among the sub-sample that experienced online racism during the pandemic, older Asian Australians reported having experienced breathing difficulty in the absence of physical exertion 'almost always': 9.3% for 45 years and over, 18.0% for 35-44 year-olds, 10.4% for 25-34 year-olds, and 3.7% for 16-24 year-olds (see Table 5). Males (12.1%) reported greater anxiety compared to females (8.7%). More Australian-born respondents (13.0%) reported more breathing difficulties compared to those born overseas (5.8%). Similarly, those who speak English at home reported greater anxiety (12.5%) compared to those who speak a language other than English at home (6.9%). Similar to the findings for stress, those who were religious reported greater anxiety compared to the non-religious; this was 15.7% for non-Christians, 8.3% for Christians, and 3.8% for those who were non-religious and who reported having experienced breathing difficulty 'almost always'. For depression among the sub-sample that experienced online racism during the pandemic, those who were aged 35-44 years of age and 16-24 years of age reported the highest proportion of feeling 'not worth much as a person' 'almost always': 15.7% for 16-24 year-olds, 8.6% for 25-34 year-olds, 22.7% for 35-44 year-olds, and 8.0% for those 45 years of age and older (see Table 6). Males reported poorer mental health, with 17.5% of males having felt not much worth as a person as compared to 9.9% of females. Similar to previous mental health indicators, 17.9% of Australian-born participants reported 'almost always' not feeling much worth as a person as compared to 7.9% of those born overseas. Likewise, in terms of religion, those who were religious reported experiencing more depression as compared to the non-religious: 16.1% for non-Christians, 15.5% for Christians, and 9.8% for the non-religious.  Table 5 presents the frequencies on the sense of belonging indicators for the 530 Asian Australians who reported having experienced online racism at least once during the COVID-19 pandemic. A little over half agreed that they felt accepted by others (52.0%), feel connected with others (53.3%), have a sense of belonging (53.3%) and feel that they are Australian (54.4%). Three in five respondents (60.9%) reported that the majority of their friends are from the same cultural background as them. Almost two-thirds (65.1%) reported feeling included when they are with other people, and almost three-fourths (73.6%) said that they have close bonds with family and friends. Almost two-thirds (65.3%) said that it is important that their children are/would be fully accepted as Australians, and approximately three-fourths-72.5%-said it is important that they continue to practice their cultural heritage.

The Relationship between Online Racism and Sense of Belonging
The relationship between experiencing online racism during the pandemic and sense of belonging was examined via cross-tabulations and chi-square tests (see Table 6). The crosstabulations show statistically significant relationships between participants' experience of online racism during the pandemic, and the sense of inclusion, bonds with family and friends, feelings of acceptance and connection with others, and aspirations for children to be fully accepted as Australians. Across these four 'belonging' indicators, Asian Australians who experience more frequent online racism (as compared to those who experienced less or no online racism but racism in other settings) reported feeling less included, accepted, and connected with others and had fewer close bonds with family and friends. Those who experienced more online racism were also more likely to agree/strongly agree that it is important that their children are/would be fully accepted as Australians.
The demographic pattern for those most likely to experience a greater sense of belonging were males, those 25-44 years of age, English speakers at home, and Christians. There were significant demographic differences in all nine indicators of belonging. Due to space constraints, only selected significant demographic patterns will be highlighted here.
When asked 'I have a sense of belonging', those aged 25-44 years 'agreed or strongly agreed' the most that they have sense of belonging: 48.5% for 16-24 year-olds; 57.3% for 25-34 year-olds; 59.1% for 35-44 year-olds; and 42.7% for those aged 45 years and above. In terms of gender, males (58.9%) were more likely to agree than females (46.3%) that they belong. Country of birth did not make much difference in terms of agreeing (53.0% for Australian-born vs. 54.2% for overseas-born); however, overseas-born respondents were more ambivalent (31.6% said neither) as compared to Australian-born respondents (19.7% said neither), and the Australian-born respondents were more likely to 'disagree or strongly disagree' that they belong (27.3% vs. 14.2% for overseas-born). A similar pattern was found for main language spoken at home, with both those who spoke English at home those who did not having similar proportions in agreement (53.3% and 53.4% respectively). However, non-English speakers were more ambivalent (32.2% said neither) as compared to English speakers (20.3% said neither), and English speakers were more likely to 'disagree or strongly disagree' that they belong (26.4% vs. 14.4% for non-English home speakers). For religion, Christians (58.7%) 'agree or strongly agree' the most that they belonged, as compared to 53.7% of those who were non-religious, and 50.0% of non-Christians.
Similar to belonging, those aged 25-44 years 'agreed or strongly agreed' the most that they feel connected to others: 46.3% for 16-24 year-olds, 58.5% for 25-34 year-olds, 59.1% for 35-44 year-olds, and 42.7% for those aged 45 years and above. For country of birth, 58.4% of overseas-born respondents felt more connected as compared to 50.6% of Australian-born respondents. However, 54.7% of respondents who spoke English at home felt connected to others, as compared to 50.6% of non-English home speakers. However, the English home speakers (23.5%) were also more likely to 'disagree or strongly disagree' that they felt connected with others as compared to non-English home speakers (17.2%). In terms of religion, Christians (58.7%) agreed the most that they felt connected to others, followed by non-Christians (51.6%) and the non-religious (50.7%). However, non-Christians (30.5%) disagreed the most that they felt connected to others, followed by 14.7% of Christians, and 11.9% of the non-religious.
When asked to consider the statement 'feel I am Australian', similar demographic patterns emerged. For example, those aged 25-44 years 'agree or strongly agree' most that they feel like they are Australian: 45.1% for 16-24 year-olds, 57.3% for 25-34 year-olds, 59.1% for 35-44 year-olds, and 54.8% of those aged 45 years and above. Most males (61.2%) felt Australian as compared to fewer than half (46.7%) of females. Furthermore, more respondents born in Australia (60.1%) felt Australian, as compared to 44.6% of those born overseas. Similarly, more respondents who spoke English at home (59.3%) felt like they were Australian as compared to those who spoke a language other than English at home (44.4%). Lastly, Christians (66.2%) were the most likely to 'agree or strongly agree' that they feel Australian, followed by non-Christians (53.5%) and the non-religious (43.6%). However, most non-Christians (30.2%) 'disagree or strongly disagree' that they feel Australian, as compared to 26.3% of the non-religious, and 18.3% of the Christians.

Discussion and Conclusions
The results from the online survey of 2003 Asian Australians not only indicate the prevalence of online racism experienced during the COVID-19 pandemic but also the potential impact of these experiences on Asian Australians' mental health, wellbeing and feelings of belonging. While the experiences of online racism (and racism across other settings) decreased during the pandemic (in contrast to some anecdotal reports), the extent of online racist experiences remains worryingly high. Of the full sample, 40 per cent (n: 798) of participants experienced racism during the COVID-19 context. Within that group, 66.4 per cent (n: 530) experienced racism online. Furthermore, the online setting was the most frequent setting in which racism occurred. Given the sharp increases in Australians' online activity since 2019 (ACMA 2021a), the empirical evidence of the prevalence of online racism, particularly online racism experienced by Asian Australians during the pandemic, is a cause for concern.
This study has also uncovered important insights into the demographic groups (within the Asian Australian cohort) most likely to experience online racism during the pandemic. Our findings indicate that Asian Australians that are young adults, males, born in Australia, English speakers at home, non-Christians, or migrants who have been in Australia less than 20 years were most likely to experience online racism since January 2020. These demographic findings in relation to age, gender, language spoken at home and birthplace (overseas vs. Australia) are consistent with Asian Australians' experiences of racism more broadly (i.e., in other settings). They are also consistent with experiential variations within the wider Australian population prior to the pandemic. For example, the 2019 Mapping Social Cohesion survey results showed that in Australia, 18-24-year-olds (31%) experienced the highest rates of discrimination, and this proportion steadily decreased with age (Markus 2019). The 2019 Mapping Social Cohesion survey results also showed that males experienced higher rates of discrimination as compared to females (20% vs. 16%). However, the survey showed a different pattern of results in relation to country of birth. Specifically, Markus (2019) found that people born in an English-speaking country (outside of Australia) reported the lowest rates of having experienced discrimination (11%), followed by 17% of people born in Australia, with people born in a non-English-speaking country having experienced the highest rates of discrimination (29%). Understanding the particular demographic variations in online racism experiences (and racism in other settings) is crucial for developing targeted anti-racism strategies, reporting tools and support services for targets and witnesses of racism. The findings reported in this paper have provided specific demographic information relating to Asian Australian experiences of racism in online settings that can be used for these purposes. However, further research is needed to understand why these particular demographic groups of Asian Australians are more likely to experience racism in online settings. Factors such as racism 'literacy' (i.e., targets recognising racism), increased exposure and presence in online settings (e.g., younger people spending more time online), and/or focus on online racist attacks/content may influence Asian Australians' online experience and thus should inform anti-racism strategies.
While the morbid effects of racism on targets and witnesses to racism have been well documented (e.g., Paradies et al. 2015), studies examining the effects of online racism are relatively sparse . The results from our study show that Asian Australians who experience frequent online racism report poorer mental health and wellbeing, as compared to those who experienced less frequent or no online racism. The relationship between frequent experiences of online racism and poorer outcomes is consistent across all the mental health and wellbeing indicators and is particularly stark when comparing the mental health of participants who did experience online racism (hardly ever-very often) with those that 'never' experienced online racism (but experienced racism in other settings). The comparison between pre-pandemic and during-pandemic experiences also suggest that online racism during the pandemic has negatively affected previously 'resilient' Asian Australians. There was an almost 15% increase in participants that found it difficult to 'snap back' during the pandemic, and this is associated with experience(s) of online racism. In other words, those who showed resilience (an ability to 'snap back') before the pandemic, may not be as resilient during the pandemic, and this is linked to their experiences of online racism. Within the sub-sample of those who experienced online racism during the pandemic, males, middle/aged or older people, those born in Australia, English speakers at home, migrants who have been in Australia for over 10 years, and non-Christians, were most likely to experience poorer mental health outcomes. This pattern of findings was consistent across the majority (11 of 14) of mental health and wellbeing indicators. Investigating what drives these particular associations between online racism and poor mental health is a matter of urgency. In addition, with mental health services often failing to capture information on cultural and linguistic diversity, which in turn limits the level of service access or mental health outcomes in some diverse communities (NMHC 2020), the mental health and wellbeing of specific groups of Asian Australians that have experienced online racism is of particular concern.
The anticipation of racism felt by the majority (71.5%) of participants who had experienced online racism during the pandemic, and avoidance of specific situations because of racism (81.2%), points to additional mental health and wellbeing stressors linked to online racism. Itaoui (2016) and Dunn and Hopkins (2016) have found that experiences and even the anticipation of racism has profound impacts on individuals' mobility and feelings of safety. This may in turn negatively impact individuals' access to essential services such as healthcare, employment, housing and education. It is also important to note the strong positive association between 'never' experiencing online racism and 'never' having to avoid or anticipate unfair treatment because of racism (see Table 4). While obvious, it is important to reiterate that those who do not experience online racism had more positive outcomes in their everyday lives, including wellbeing and inclusion, indicating an anti-racism position to strive towards.
The survey results have also indicated the relationship between experiencing online racism and a reduced sense of belonging. Asian Australians who experience more frequent online racism (as compared to those who experienced less) reported feeling less included, accepted, and connected with others and had fewer close bonds with family and friends. The demographic pattern of those most likely to experience a greater sense of belonging (who had experienced online racism) were males, those 25-44 years of age, English speakers at home, and Christians. There were significant demographic differences in all nine belonging indicators. It is also important to reiterate that those that did not experience any online racism (but experienced racism in other settings) were more likely to report a sense of belonging, and this was statistically significant (Table 6). This points to the particularly corrosive nature of online racism on individual wellbeing as well as social cohesion. During a time of social crisis (such as the COVID-19 pandemic), when a sense of community, cooperation and trust are essential in overcoming the associated challenges, a deterioration in the sense of belonging among particular groups in society is problematic (Quinn et al. 2021).
Finally, in the majority of cases, online racist incidents experienced by Asian Australians were directly linked to the COVID-19 context. This supports the findings of Kassam and Hsu (2021) who also found that the majority (66%) of racist incidents experienced by Chinese Australian survey participants were COVID-19 related. Therefore, while our survey found a decrease in racism experienced by Asian Australians (online and more broadly), we have provided empirical evidence for the predominance of COVID-19 in relation to online racism. This illustrates the far-reaching impacts of the racialisation of the pandemic-in media, public and political discourse-on the everyday 'online' (and offline) experiences of Asian Australians.