More than Money: Examining the Potential Exposure of Romance Fraud Victims to Identity Crime

ABSTRACT

Romance fraud occurs when an offender uses the guise of a genuine relationship to defraud an individual for financial gain. Known statistics indicate that millions are lost each year to this form of fraud. However, the potential for romance fraud victims to also experience identity crime as a consequence of their interactions and communications with an offender has yet to be explored. This article begins to address this gap by examining the potential exposure for identity crime in the context of romance fraud using reports lodged with Scamwatch, an Australian online fraud reporting portal. The findings demonstrate few behavioural and demographic characteristics associated with the potential exposure of those targeted by romance fraud to identity crime. However, the article provides insights into how this analysis should inform prevention messaging and victim support services for those targeted by romance fraud.

KEYWORDS:

• Romance fraud
• identity crime
• identity theft
• scams
• cybercrime, victims

Introduction

Romance fraud occurs in ‘situations where an individual is deceived for financial gain by someone with whom the victim perceives to be in a romance relationship’ (Buil-Gil & Zeng, 2021, p. 1). In this scheme, victims voluntarily transfer funds to the offender through various electronic methods under the guise of assisting their partner. The losses associated with romance fraud are significant across the globe. For example, in 2020 over USD$600 million was lost in the United States of America (, 2021, p. 20), £68 million was lost in the United Kingdom (Wakefield, 2021), and C$18.5 million was reported lost in Canada (Fraud Centre CAFC, 2021), Losses are not restricted to the Global North, with Hong Kong reporting that HKD$160.8 million was lost to victims in 2020 (Lee, 2020), with a doubling of victims in the first half of 2021 compared to 2020 (Heung & Yau, 2021). Additionally, countries such as Malaysia (Othman, 2020) and South Africa (SABRIC, n.d.) have also put forward warnings to citizens about the potential for this crime type. Of notable concern, the current COVID-19 pandemic has seen a substantial increase in romance fraud cases, likely attributed to the far ranging and severely disruptive changes witnessed during the pandemic (Buil-Gil & Zeng, 2021).

Australia is not immune to this, with victims having reported losses of AUD$131 million in 2020, up from AUD$83 million the previous year (Australian Competition and Consumer Commission ACCC, 2021; 2020). These figures likely represent a small portion of the actual losses to romance fraud as victims are known to have low reporting rates and in addition, such statistics do not account for the many non-financial harms experienced (Button & Cross, 2017). In fact, romance fraud victims have an extra burden of what is termed the ‘double hit’ of victimisation, (M. Whitty & Buchanan, 2012), whereby victims must grieve the loss of the relationship as well as any monetary losses they may have sustained (Cross et al., 2016).

To date, there is limited research examining romance fraud using quantitative data (M. T. Whitty, 2018). Studies typically utilise qualitative data and focus on those who have experienced financial losses, and the associated impacts of victimisation across various aspects of their lives (e.g. Cross et al., 2016; M. Whitty & Buchanan, 2012). There are also several studies that seek to determine indictors for potential victimisation (Buchanan & Whitty, 2014; M. T. Whitty, 2018). Though beneficial, there has been limited research exploring the experiences of those targeted by romance fraud, but who have not sustained direct financial losses. These individuals can still sustain significant trauma as a result of their experience.

There is a dearth of research that examines the interactions between those targeted by romance fraud and other crime types, such as identity crime. Though victims of romance fraud directly provide their fraudster with financial support, they may also give out sensitive personal information, such as bank account details, as a part of their communication with an offender (Cross et al., 2016; M. Whitty & Buchanan, 2012). There is also emerging evidence to suggest that some victims experience a sustained fear of (identity) crime from having shared personal details with an offender (Cross & Lee, 2022).

It is unclear what personal and situation-specific factors may increase the risk of financial harm from identity crime for these victims, outside of any monetary losses suffered as part of romance fraud victimisation (usually through the direct transfer of funds). Thus, this article seeks to address this gap in the literature by investigating any demographic and behavioural factors associated with the loss of banking details and financial losses, both individually and combined, for those targeted by romance fraud, using data provided by complainants to Scamwatch (an Australian online fraud reporting platform).

Specifically, this analysis estimated three binary logistic regression models to assess the relationship between behavioural and personal characteristics and the risk of experiencing the loss of personal details, banking details, and financial losses. The implications of these analyses for an understanding of romance fraud related to the prevention and support of victims are examined in depth.

Understanding romance fraud

Romance fraud ‘is a method used by offenders whereby they use the perception of a legitimate relationship in order to manipulate and exploit the victim’ (Cross, 2020, p. 3). Offenders will seek to use the guise of a genuine relationship, developing trust and rapport with the victim through constant communication across various channels, including email, text messaging, phone calls, and video calls (Cross et al., 2016).

There are many well-known impacts of romance fraud which can be devastating on an individual. Apart from the obvious monetary losses noted in the introduction, victims often experience a range of non-financial harms that range from a deterioration in physical and emotional wellbeing, varying levels of depression, relationship breakdown, unemployment, homelessness and in extreme cases, suicide (Button & Cross, 2017; Button et al., 2009; Cross et al., 2016). Romance fraud also carries the ‘double hit’ of victimisation (M. Whitty & Buchanan, 2012), whereby victims must not only grieve the loss of any money, but they also grieve the loss of the relationship. Coluccia et al. (2020, p. 24) also note that victims also endure a large degree of shame upon discovering they have been defrauded, which impacts on accurate reporting rates. The negative victim stereotypes and the stigma of fraud victimisation is a significant additional barrier that many romance fraud victims must navigate (Cross, 2015, 2018a).

It is well established that fraud has low rates of reporting (Button et al., 2014; Copes et al., 2001; van Wyk & Mason, 2001), therefore any of the known statistics of romance fraud are likely to severely underestimate an accurate measure of financial losses incurred as a result of this crime type. These official figures also fail to take into account the non-financial harms experienced by victims (Cross, 2020).

It is also difficult to establish strong romance fraud victim profiles, given that vulnerability to fraud cuts across demographics and other factors (Cross, 2020). There has been a number of psychological studies which have sought to explore the factors associated with victimisation. For example, M. T. Whitty (2018) examined the profiles of romance fraud victims compared to non-victims against a number of variables, including age, gender, education knowledge about cybersecurity, impulsivity, locus of control, trust in others (gullibility), trustworthy individual, kind, greed and addictive disposition. Results of this analysis indicated that victims more likely to be middle aged, have higher education levels, possess traits that measured impulsivity and lack of self-control, and be less kind (M. T. Whitty, 2018). In an Australian context, known statistics on romance fraud victims support some of these assertions. For example, women experience higher financial losses from romance fraud compared to men, and those aged 45–64 are most likely to report romance fraud victimisation (Australian Competition and Consumer Commission ACCC, 2020).

Linking identity crime to romance fraud

While these studies and figures give some insights into who becomes a victim, there is still work to be done in this area. In particular, there is a need to understand the degree to which pieces of personally identifiable information are essential to the fraudster in order to experience financial gain, as that is primary goal of romance fraud in general (Cross, 2020; Fletcher, 2007). Personal information is a valuable commodity that may be provided to offenders, which creates additional opportunities for offenders to access victim bank accounts and lines of credit (both existing and new) to commit identity crimes.

As a term, identity crime can be understood as an ‘umbrella concept’ (Wall, 2013, p. 437) that includes ‘identity theft, identity fraud, identity fabrication, identity manipulation, [and] lent identities’ (Smith, 2011, p. 142). Identity crime encompasses ‘crimes that are in some way enabled by the use of an identity that is not the identity of the perpetrator’ (Steel, 2010, p. 48). Identity theft can be defined as ‘the unlawful use of another’s personal identifying information’ (Allison et al., 2005, p. 19) with personal information capturing a variety of variables such as “name, address, driver’s licence, social security number, telephone number, place of employment, employee identification number, mother’s maiden name, demand deposit number, saving account number, or credit card number (Milne, 2003, pp. 390–392).

In this regard, romance fraud could serve as a context for other forms of offending. In fact, Furnell (2019, p. 6) argued that cybercrime should not be viewed as a crime in isolation; instead ‘it often feeds through to other forms of criminal activity’. Holt and Turner (2012, p. 312) hypothesised that ‘individuals experiencing one or more forms of cybercrime victimisation may also face an increased risk of identity theft’. Consequently, it is important to explore whether there is a relationship between the two forms of cybercrime victimisation, romance fraud and identity crime, and whether there is a relationship between financial losses and loss of identity information.

Obtaining identity information is an attractive method for offenders to gain financial advantage. Offenders can use sensitive personal details for a variety of reasons ‘including accessing existing credit cards or bank accounts without authorisation, opening new credit card accounts, and other forms of fraud’ (Holt & Turner, 2012, p. 308). Further, ‘once an identity thief is in possession of a victim’s identity information there are a number of ways in which it may be used to gain cash or to convince the credit card agency to issue a duplicate card on an existing account’ (Vieraitis et al., 2015, p. 12). The offender does not have to persuade the victim to send money directly, or alternatively, can employ their information if/when a victim has refused to send (continuing) amounts of money in response to a request.

The establishment of an online profile and communication between a victim and offender enables the sharing of intimate personal details, particularly during the initial stages of the relationship. It is difficult to effectively engage in social interactions on dating platforms without disclosing elements of a person’s identity (Winterdyk & Filipuzzi, 2009, p. 309). To protect oneself, an individual can ‘control the release of information about themselves online’ (Burns & Roberts, 2013, p. 50). In the context of an online relationship this is both logistically and ideologically difficult and, to some degree, contradicts the purpose of online dating. Participants ideally seek to establish a connection with someone and to build trust and rapport through open communication. This provides ideal grounds for romance fraud offenders to gather information about their victims. Vieraitis et al. (2015, p. 12) noted that ‘some offenders create elaborate schemes to dupe victims into revealing their personal information both on and offline’, and again, this is arguably relevant to romance fraud.

The current study explored whether there is evidence to suggest any relationship between those targeted by romance fraud and identity crimes, particularly the experience of a loss of personal information, and financial losses. A key question guided this analysis: what are the demographic and behavioural factors associated with the loss of banking details and financial losses individually, and both forms of loss combined for those targeted by romance fraud? A set of three binary logistic regression models were estimated using a quantitative data set of self-reported victims from Australia. Multiple behavioural and demographic variables were included to explore the predictors for self-reported losses of personal information, banking information, and actual financial losses.

Data and methods

The data for this analysis was obtained from the Australian Consumer and Competition Commission (ACCC) of reports made through their Scamwatch portal. A written request was submitted to the ACCC, who released a deidentified excel spreadsheet of all romance fraud reports received between July 2018 and July 2019 (inclusive) (see also Cross & Holt, 2021; Cross & Layt, 2021; Cross & Lee, 2022 for other studies to use the same dataset). An ethics exemption was granted by Queensland University of Technology Human Research Ethics Committee (#1900000738) for this data request.

As noted earlier, Scamwatch is an online reporting mechanism for fraud within Australia. While targeted at Australian victims, the online nature of the portal means that victims across the globe can lodge a report on an incident which may or may not have an alleged Australian link. There were 4,354 reports filed during this timeframe, with 3,463 (80%) reports indicating that the complainant was willing to share the contents of their report for scam awareness. Scamwatch is not a law enforcement agency and cannot investigate or take action regarding individual fraud complaints. Rather, the information gathered by Scamwatch is used for intelligence purposes to drive consumer awareness and education across Australia. It also drives a range of disruption activities undertaken by the ACCC with other agencies (Australian Competition and Consumer Commission ACCC, 2021).

The self-reported nature of this data limits its generalisability as Scamwatch is an online reporting mechanism with no screening or human interaction in the reporting process. It is similar in structure to the US Internet Crime Complaint Center’s reporting mechanism which creates issues for victims and complainants when trying to provide their information (see Holt & Bossler, 2015). Those reporting may have found the website through internet searches or in some cases, it is evident they were directed to report by third parties, such as police and banks. As a result, the experiences reported by complainants may not be reflective of the experiences of all romance fraud victims generally.

Further, the self-report nature of this data limits its potential reliability, as responses were not mandatory to all questions presented. In addition, Scamwatch provided limited definitions and guidance to respondents while completing all online forms. Complainants’ interpretation of phrasing may hinder their potential accuracy when replying. Finally, the agency took no action to verify any of the information provided by complainants. The data is taken at face value and is limited by the extent to which the respondent can accurately recall the events as they occurred.

Despite these challenges, the data provides some important insights into the relatively hidden aspects of romance fraud victimisation, and in this case, how it relates to potential exposure to identity crime. The difficulties inherent in accessing fraud victim populations (see Cross, 2020), coupled with the general absence of quantitative data sources to assess victimisation and experiences with fraud (Holt & Bossler, 2015) make this data invaluable to understand the contours of various fraud schemes, and identity crime victimisation in particular.

The ACCC provided a de-identified excel spreadsheet containing the details of these 3,463 reports. Each report included the following information: demographic details of the complainant (gender, age, and jurisdiction both within Australia and overseas), details about the fraud (how the approach was received, the location of the alleged offender), and any losses incurred (amount, payment methods, sensitive details lost). Vulnerability indicators of the complainant were also included, particularly around age, disability, sickness, financial hardship, and location. The de-identified free text field where each complainant wrote a summary/description of what happened was also made available to the researchers.

In reviewing the data, 204 duplicate entries were removed, leaving 3,259 distinct reports available for final analysis. An additional 253 cases were removed as they involved sextortion complaints which may differ from instances involving online dating or related forms of fraud. A further 307 cases were removed due to missing data and related issues. As a result, this article focuses on an analysis of 2,699 reports which is a substantive reduction in the total population. At the same time, the final sample is still relatively reflective of the full sample population on various key demographic variables including gender (female = 56.8% full; 57.5% final), Australian nationals (63.7% full; 66.1% final) and financial hardship status (15.9% full; 17.2% final). Thus, the data should be generalisable to the broader population of complainants.

Dependent variable

This analysis focuses on three dependent variables. The first was the reported loss of bank details (9.8% all complaints; see Table 1). There was no additional information provided to complainants so that they could better understand what these terms meant. As a result, the self-report nature of these measures should be viewed with caution.

Table 1. Descriptive Statistics of Quantitative Measures (n = 2,699).

Variable	Mean	S.D.	Min	Max
Financial loss	.339	.473	0	1
Banking details loss	.098	.298	0	1
Banking and financial	.058	.234	0	1
Social networks	.339	.473	0	1
Personal details	.316	.465	0	1
International	.308	.462	0	1
Disability status.100	.300	0	1
ESL	.084	.277	0	1
Financial hardship	.172	.377	0	1
Remote community	.049	.217	0	1
Illness	.040	.196	0	1
Indigenous	.031	.173	0	1
Age	3.812	1.469	1	6
Female	.575	.494	0	1

The second binary variable for financial loss (0=no; 1=yes; 33.9%) was created to reflect those complainants who either noted in a checkbox that they suffered a financial loss or indicated in the free text section of the report that they had sent money to the offender(s). Reading the free text section was essential to determine monetary loss, as well as gain a more comprehensive understanding of who sent money as a result of a specific fraudulent approach. Since the ACCC was not able to verify all reported complainant losses, a binary variable was used in lieu of the reported dollar figure. Such a conservative measure allows for an assessment of factors associated with monetary loss without being impacted by complainant over or under-reporting of dollar value.

The third variable recognises those who reported experiencing both forms of victimisation (banking and financial; 0=no; 1=yes). Only 5.8% of the sample experienced both forms of victimisation, suggesting that individuals who experienced one form did not necessarily experience both. Thus, a model was estimated to identify any differences in the correlates of loss.

Independent variables

When filing a complaint, respondents were also presented with various drop-down menus and options related to both their personal demographic characteristics and the nature of the offence. For instance, complainants could identify the way in which they were initially contacted by the offender, including: 1) email; 2) fax; 3) internet; 4) in-person; 5) mail; 6) mobile apps; 7) phone (voice); 8) social networking/online forums; or 9) text message. Only one category could be selected by a complainant, and the majority identified being contacted via social networking sites or forums. Thus, a binary measure was created to reflect contact via social networks.

Respondents were also asked to identify what sorts of information they lost in the course of the offence. The reported loss of personal details (31.6% all complaints) was included to assess any connection between the loss of sensitive information and fraud. There was no additional information provided to complainants so that they could better understand what this term meant. In reviewing the free text narratives, it is clear that some complainants gave away only their names and email addresses, while others sent copies of their driver’s licence and/or passport. In this way, the category should be understood a broad in scope and as with previous variables, the self-report nature of this measure should be viewed with caution.

Complainants were also asked whether they resided in Australia or an international jurisdiction (0=Aus; 1=international). Additional demographic variables were captured in reports to better understand the extent to which victims may fall into protected or vulnerable populations. Specifically, complainants were presented with a series of checkboxes to indicate their: 1) disability status (10%); whether English is their second language (ESL; 8.4%); if they are living in financial hardship (17.2%); reside in a remote community (4.9%); have a chronic/serious illness (4%); and identify with Indigenous status (3.1%). Similar to the loss variables noted above, no additional information was provided to respondents to explain these terms or authenticate their claims. While it is hoped that any complainant would correctly identify that they fall into one of these categories, there is the potential they may have been selected in error or that these were interpreted differently across the sample. As a result, they should be treated with caution.

Two additional control variables were included in the analyses. First, a categorical variable for complainant age was included, as respondents were presented with seven responses in the original dataset: 1= under 18; 2 = 18–24; 3 = 25–34; 4 = 35–44; 5 = 45–54; 6 = 55–64; 7 = 65 and over. This item was recoded into a six-item measure to exclude those under 18 years of age (n = 23), given the current analysis is focused on adult experiences of victimisation. Second a binary variable was created to reflect complainant gender (female: 0=male; 1=female). Complainants were also given the opportunity to not report a gender in the original dataset, so those 52 cases were excluded from the data presented below for the sake of parity in the quantitative analysis (noting that these excluded cases relating to age and gender were removed in the initial data cleansing process outlined previously in the methodology and sit outside the 2,699 cases in the current analysis).

Results

Binary logistic regression models were estimated to assess any factors associated with reporting loss of banking details, financial losses and experiencing both in the course of reported romance fraud. There were no issues observed with respect to multicollinearity as no VIF was higher than 1.095, and no tolerance was lower than .912. The same variables were included in the models for the loss of banking details and both forms of victimisation, though the loss of banking details was included as an independent variable in the model for financial loss. This was conducted to assess any relationship between these two variables.

The results for banking details provided limited evidence of risk factors for fraud victimisation (see Table 2). Those who lost their personal details (Exp(B) = 4.861; sig=.001) and those who have been contacted via other forms of online communication than social media sites (Exp(B) = 0.735; sig=.01) were more likely to become a victim. In terms of personal risk factors, those who reported a financial hardship status were more likely to report losing their banking details (ExpB = 2.890; sig=.001).

Table 2. Binary Logistic Regression Model Estimating Correlates of Reported Loss of Banking Details by Complainants (n = 2,699).

Variables	b	S.E.	Exp(B)
Social networks	−.308	.148	.735*
Personal details	1.581	.142	4.861***
International	−.264	.153	.768
Disability status	.297	.214	1.346
ESL	−.242	.253	.785
Financial hardship	1.061	.152	2.890***
Remote community	−.085	.334	.919
Illness	−.144	.324	.866
Indigenous	.464	.341	1.591
Age	−.093	.048	.911
Female	−.278	.143	.757
Constant	−2.528	.219	.080***

Note: −2LL = 1538.708; Chi square = 198.889, Nagelkerke R2=.150; *=.05; **=.01; ***=.001

The results for financial losses demonstrated that being contacted via a social networking site was significant (Exp(B) = 1.257; sig=.01; see Table 3). Second, those who lost their bank details to offenders were significantly more likely to experience financial losses (Exp(B) = 5.258; sig=.001). Those who did not lose their personal details were also more likely to experience financial losses (Exp(B)=.346; sig=.001). This finding may reflect the substantive value of an individual’s banking information for an offender, relative to personal details like address, maiden name, or date of birth. It is far easier to cause immediate financial harm with credit or debit card information through fraudulent transactions (see also Holt & Bossler, 2015; Newman & Clarke, 2003).

Table 3. Binary Logistic Regression Model Estimating Correlates of Reported Financial Loss by Complainants (n = 2,699).

Variables	b	S.E.	Exp(B)
Social networks	.229	.094	1.257*
Banking details	1.387	.149	5.258***
Personal details	−.990	.109	.346***
International	−.328	.099	.964***
Disability status	−.312	.151	.732*
ESL	.064	.164	.980
Financial hardship	1.353	.113	3.719***
Remote community	−.078	.207	.857
Illness	−.256	.232	1.010
Indigenous	.006	.255	.952
Age	−.027	.031	1.010
Female	−.388	.091	.702***
Constant	−.413	.142	.734**

Note:-2LL = 3109.664; Chi square = 349.749***, Nagelkerke R2=.168; *=.05; **=.01; ***=.001.

Those victims living in Australia were more likely to report financial losses (Exp(B)=.964; sig=.001), as were those who did not identify with a disability (Exp(B)=.732; sig=.01), and living in financial hardship (Exp(B) = 3.719; sig=.001) in keeping with prior research that those in lower socio-economic status groups were more likely to experience identity theft (Harrell, 2019). Male complainants were also significantly more likely to experience financial loss (Exp(B)=.702; sig=.001), in keeping with prior research that found higher rates of economic loss reported by males (e.g. Button & Cross, 2017; Harrell, 2019).

The model estimating the correlates for the loss of both banking details and funds demonstrated consistent correlates of victimisation relative to the prior models (see Table 4). Those who lost their personal details were more likely to report experiencing both forms of victimisation (Exp(B) = 5.105; sig=.001). In terms of personal risk factors, those who lived in Australia (Exp(B)=.571; sig=.01), and reported living in financial hardship (Exp(B) = 3.524; sig=.001) were more likely to report losing both their banking details and experienced financial loss.

Table 4. Binary Logistic Regression Model Estimating Correlates of Reported Loss of Banking Details and Funds by Complainants (n = 2,699).

Variables	b	S.E.	Exp(B)
Social networks	−.113	.183	.893
Personal details	1.630	.182	5.105***
International	−.560	.203	.571**
Disability status	.432	.262	1.540
ESL	−.259	.329	.772
Financial hardship	1.260	.182	3.524***
Remote community	−.185	.452	.813
Illness	−.473	.432	.623
Indigenous	.253	.459	1.288
Age	−.107	.062	.898
Female	−.148	.183	.863
Constant	−3.238	.283	.039***

Note:-2LL = 1049.688; Chi square = 148.136, Nagelkerke R2=.149; *=.05; **=.01; ***=.001.

Discussion

It is evident that romance fraud victims can potentially provide a large amount of personal information (including financial details) to offenders, which may expose them to identity crime. The current analysis utilised a quantitative sample of self-reports made by individuals who were targeted by romance fraud, and explored the factors associated with individuals’ reports of financial losses associated with romance fraud. Finally, the factors associated with both the loss of banking details and financial losses were explored through a binary regression model.

The above analysis indicates a clear relationship between the loss of personal information and victims’ reporting both a loss of banking details and experiencing economic loss. While the loss of personal information was negatively associated with financial losses on their own, it was positively associated with both the loss of banking details alone, and both forms of victimisation when combined. Access to sensitive information about a romance fraud victim may enable offenders to engage in various forms of identity crime and may be an attractive option for offenders to more effectively request funds during their relationship with the victim. Retaining this information for use in identity crimes may also present an alternative for offenders when victims cease complying with any financial requests. Victims’ willingness to share information is also likely to be a by-product of the trust and rapport they have developed with the offender, and an indicator of the seriousness of their relationship. This maps to the stages set out in the ‘scammer’s persuasive techniques model’ (M. Whitty, 2013), and affirms that the sharing details could legitimately form part of the grooming process and a prelude to fraudulent transactions.

Individuals who reported financial hardships were also more likely to report both the loss of banking details and money to the offenders. It is unclear why such a dynamic is present, as prior research has found mixed evidence of individuals in both higher and lower socioeconomic groups reporting identity theft losses (Burnes et al., 2020; Harrell, 2019; Reyns, 2013; Williams, 2016). It may be that individuals experiencing financial hardships may be more inclined to respond to solicitations that may improve their economic outlook. However, the structure of the reporting form within Scamwatch may have created a temporal ordering issue, in that individuals who filed complaints of romance fraud losses are facing financial hardship as a result of their victimisation. Further research is needed with victim populations using both qualitative and quantitative measures to disentangle this relationship and determine if the financial hardship was in existence prior to the targeting of romance fraud or occurred in the aftermath of an incident.

This study also found minimal support for the relationship between victim demographic characteristics and their risk of reporting identity fraud and theft as a result of romance fraud. This mirrors previous romance fraud research which has struggled to find demographics as a meaningful indicator of romance fraud victimisation as a whole (Button & Cross, 2017; M. T. Whitty, 2018). Overall, these findings further reinforce prior research indicating that romance fraud victims may come from any background (ACCC, 2019), and that demographics alone are not a strong indicator of differentiating victims from non-victims.

Conclusion

As society becomes more dependent on technology for all aspects of communication and personal finance, the opportunities for economic fraud offences have increased dramatically (Button & Cross, 2017; Newman & Clarke, 2003). Research on various forms of online fraud have grown over the past decade, with an increasing emphasis on romance fraud schemes due to their emotional and financial impact on victims (Buchanan & Whitty, 2014; Button & Cross, 2017; Cross et al., 2018; M. Whitty, 2013). To date, many studies have focused on seeking to predict factors for victimisation, as well as documenting the ways offenders perpetrate romance fraud and the severe impacts of victimisation.

Taken as a whole, this study demonstrates the significant relationship between loss of personal information and the risk of experiencing both the loss of banking information and experiencing financial harm in the context of romance fraud. Current prevention messaging on romance fraud focuses heavily on both the request and subsequent transfer of money, as well as banking and/or credit card details. Programming should be added related to the need to guard against the release of personal information more broadly, as the seemingly innocuous questions that an offender may ask in order to gain a wholistic picture of their victim, could be used a means to facilitate identity crimes at a later date.

To that end, the diffuse nature of romance fraud victimisation suggests targeted messaging campaigns to demographic groups may have minimal effect in reducing harm. A lack of demographic variables to determine victims versus non-victims more broadly, combined with the results of the current analysis, point to the potential of a greater benefit in presenting carefully crafted messages to users on social media and other online platforms that highlight the risk of sharing personal information with others. Users appear to still be largely unaware of the value of their personal details, and the ways that offenders can use these details for their own financial benefits, particularly within the context of romance fraud.

Therefore, prevention messaging should focus more on the action of disclosing sensitive information, rather than attempting to focus on the individual themselves and factors relating to them personally. As part of this, it would also be important to highlight the ways that personal details can be used as financial currency in the same way that money itself can be used. These messages appear in a more generalised sense when warning against identity crime but are yet to be specifically connected to romance fraud, and those seeking to pursue online relationships.

Prevention materials (including videos, images, and text-based narratives) should emphasise the risk of identity fraud and the logistics of how this is perpetrated stemming from what may be viewed as an innocent behaviour. This may capture individuals’ attention and help inform them of the risks of an identity crime (Harrell, 2019; Newman & Clarke, 2003). Such efforts may be instrumental in reducing victims’ willingness to provide sensitive data to others and improve their ability to recognise fraudulent schemes at early stages of engagement (M. Whitty, 2013). The use of victim stories to highlight the realities of victimisation and the potential for identity crime may be a useful mechanism to increase awareness to those engaged in online dating practices. It may also seek to reinforce the need to promote critical thinking and scepticism as it relates to online communication, with ‘many people are generally naïve believers of information they receive especially at the initial stage of communication’ (Jimoh & Stephen, 2018, p. 41).

These results also provide emphasise the ongoing need to provide better support to those who are both targeted for romance fraud, and sustain monetary losses from romance fraud. There are currently limited support services available to all fraud victims (Button & Cross, 2017; Cross, 2018b). The current analysis highlights the need to engage those targeted by romance fraud and who have disclosed a range of details to an offender to take steps to protect their identities from future harm. The types of actions recommended to these individuals may mirror advice received by those who have experienced a data breach, such as notifying banks and taking out a credit history report (Burnes et al., 2020). This may reduce the success of an offender using these details in the future to perpetrate further criminal acts. Within Australia, iDcare are a victim support centre for individuals who experience identity crime (idcare.org). Increased promotion and referrals to iDcare (at least in an Australian context) may be of benefit to those concerned to reduce the likelihood of potential incidents.

The continued increasing statistics related to both romance fraud and identity crime indicates a pressing need for more effective prevention and disruption techniques to avoid the devastating losses incurred through victimisation. Both are significant in and of themselves, but combined, can be life-changing and hugely detrimental to individuals. They pose an ongoing challenge that is unlikely to reduce in size any time into the near future. The results of the current study provide an entry point to better understand the dynamics between the two and highlight the need for further research that seeks to understand the complexities and interplay between romance fraud and other offence types, including that of identity crime.

Though this study provides important insights into the unique nature of romance fraud victimisation, there are several limitations of the data that must be considered. First, the self-reported nature of the data and its primarily Australian respondent population limits its generalisability to other nations. Further study is needed with more comprehensive international sample populations in order to better understand the factors associated with the loss of personal information and financial losses associated with romance fraud.

Second, the data is focused on those who self-disclose romance fraud victimisation, which may not be generalisable to the broader population of victims who do not report their experiences to law enforcement, financial institutions, or other regulatory bodies. Thus, the findings may not be applicable to the broader population of romance fraud victims. Finally, the self-report nature of the data does not allow the researchers to identify the rationale for victims’ information sharing to the offender. Future research is needed to better understand the context for information sharing within a romance fraud scheme, and its tie to later experiences with financial loss and potential for identity crime victimisation. Such efforts can greatly improve our understanding of romance fraud and online crime as a whole.

Acknowledgments

The work has been supported by the Cyber Security Research Centre Limited whose activities are partially funded by the Australian Government’s Cooperative Research Centres Programme.

The authors would also like to thank the Australian Competition and Consumer Commission (ACCC) for their provision of data and their ongoing support of research in this area. The views expressed in this article are those of the authors alone, and do not necessarily represent those of the Australian Government. All errors and omissions are the sole responsibility of the authors.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The work was supported by the Cyber Security Cooperative Research Centre .