Who Became Victims of Financial Frauds during the COVID-19 Pandemic in Japan?

Abstract

The COVID-19 pandemic has provided a unique opportunity for fraudsters to innovatively swindle money through the trade of necessary goods and services. Although several incidents of financial fraud were reported during the pandemic, there is a lack of studies comparing financial frauds before and during the pandemic and the risk factors associated with frauds. This study uses two waves of a panel survey conducted before and during the pandemic and applies mean comparison tests and logit regressions to investigate financial frauds at the aggregate and specific levels. The comparative analysis shows no significant change in financial frauds at the aggregate level between before and during the pandemic. However, refund frauds for men have increased, while loan guarantee frauds for women have decreased significantly during the pandemic. The regression results show that being male, younger in age, living with family, having employment status, having a household income, household assets, having financial literacy, having a myopic view of the future, and having careful buying habits are associated with the probability of being victims of financial frauds during the pandemic. The study reveals differences in risk factors associated with victims of financial frauds at the aggregate and specific levels. The results further imply that risk factors differ across the types of fraud, which authorities should consider while combating financial frauds.

1. Introduction

The COVID-19 pandemic has not only disrupted the social and economic lives of people but has also created a situation where fraudsters have been using innovative swindling mechanisms to deceive people. Several reports show that fraudsters used a lack of communication, informational fictions, and health and financial concerns to develop their innovative swindling methods and target many people [1,2,3,4,5,6]. For example, fraudsters devised new strategies for phone fraud and email phishing by pretending to be hospital staff, demanding money for ailing relatives, or asking for personal information by pretending to be national or global health authorities [7]. Furthermore, we observe an increasing number of new types of financial fraud during the pandemic, including cyber fraud, telephone fraud, payment fraud, billing fraud, refund fraud, insurance fraud, and others [1,4,5,6,7,8]. However, existing research has paid little attention to the comparison of financial frauds before and during the pandemic to understand new episodes of financial fraud during the pandemic and to identify possible victims so that law enforcement agencies can design effective anti-fraud measures. This study aims to compare the magnitude of financial frauds before and during the pandemic and to identify risk factors that made people victims of financial fraud during the pandemic.

Several reasons have caused the proliferation of financial frauds during the pandemic. First, online trading has flourished as people are increasingly staying at home, working remotely, observing social distancing, and using mobile devices for online shopping and banking [9]. The increasing use of online platforms for trade and commerce has resulted in several incidents of online trading fraud during the pandemic [8]. The United States Federal Trade Commission reported that online shopping fraud was the largest fraud reported during the pandemic [10]. Online frauds were also observed in online ticketing and vacation- and travel-related transactions when travel rebounded due to COVID−19-related vaccination [10]. Second, the pandemic has dramatically changed the relative demand for goods and services; many goods and services are more in demand now than earlier, while others are less in demand. This large shift in relative demand is likely to cause significant friction and fraudulent activities [11]. Interpol [7] reported cases of fraud in medical supplies, which were in great demand during the pandemic. Many consumers complained that they were charged extraordinarily high prices for products, such as hand sanitizer, toilet paper, and masks, on opportunistic online platforms [10]. Moreover, unscrupulous elements even manipulated the stock market by providing fake information on the effectiveness of products and materials used to combat the pandemic [12]. This artificial increase in stock prices resulted in financial losses for many investors. Third, the economic shutdown and consequent loss of income and employment drove many people to engage in fraudulent activities [5]. Fourth, the scammers devised a mechanism to swindle money from the recipients of economic stimulus packages announced by the government during the pandemic [4]. Several fake agencies emerged during the pandemic that sought people’s financial information for the distribution of stimulus money.

Despite being a safe country, Japan has observed increasing evidence of financial frauds involving innovative swindling methods in recent years [2,13]. The most common financial fraud observed in Japan includes telephone frauds (“it’s me” fraud), fictitious billing frauds, loan guarantee frauds, and refund frauds [2,13,14,15]. Generally, financial fraud involves deceiving people with promises of goods, services, or financial benefits, which are either nonexistent or never intended to be delivered [16]. “It’s me” fraud, called “Ore Ore Sagi” in Japanese, is a mechanism through which fraudsters pretend to be a member of the target family and ask for money for emergency purposes. Fictitious billing frauds involve sending a fake invoice, sending the original invoice multiple times, altering the original invoice, or changing the billing address. Meanwhile, loan guarantee frauds involve collecting an upfront fee or sensitive information through a fake guarantee for a loan or credit card sanction. Finally, refund frauds involve fraudsters pretending to be staff in charge of tax, insurance, or other expenses of people and instructing them to transfer money to a designated account to receive a refund that does not exist. Although there are instances of larger financial scams like Ponzi schemes and corporate scams, these four types of financial frauds mostly affect common people [2,13,14]. How fraudsters change the swindling mechanism and whom they are going to target during the pandemic is a matter of concern. After all, scammers always take advantage of disasters and disorders to swindle money from people’s pockets.

The scenario of financial frauds during the COVID-19 pandemic in Japan is similar to that observed worldwide. Online frauds increased dramatically during the pandemic due to the changing lifestyle that involved reliance on online trading platforms and cashless transactions [17]. Many cases of fraud through fake websites claiming to supply necessary products were also evident in Japan. The National Police Agency of Japan [17] reported that the incidence of online banking frauds, digital payment frauds, fake online trading platforms, and deception through scam websites selling necessary products, such as masks and vaccines, increased significantly during the pandemic. In addition, fraudsters devised new swindling techniques using pandemic-related necessities. For example, in a case reported by Nippon.com [18], fraudsters targeted older people pretending to be government employees and asking for bank books and other credentials to help such people receive government-declared stimulus money. Although the instances of financial frauds during the pandemic have been covered in reports of newspapers and government agencies, there is a lack of a comprehensive academic study that investigates the changing nature of financial frauds and risk factors associated with fraud victims during the pandemic in Japan. This study will help to achieve financial sustainability for consumers by protecting them from being victims of financial fraud, which causes them a huge financial loss and sometimes deters them from making a profitable investment. From this point of view, the study is related to Sustainable Development Goal number 8, declared by the United Nations, which deals with promoting inclusive and sustainable economic growth, employment, and decent work for all. The disruption of economic activities during the COVID-19 pandemic was triggered in part by the increase in fraudulent activities related to the goods and services required to combat the pandemic. This study sought to find risk factors and possible victims so that smooth economic activities support economic growth.

To fill this gap in the literature, this study compares financial frauds experienced before and during the pandemic in Japan to identify changes in such frauds and risk factors associated with new victims of financial frauds at the aggregate and specific levels. This study contributes to the existing literature in at least two ways. First, it identifies the socioeconomic background of victims of financial frauds during the pandemic. Second, it identifies emerging types of financial fraud, which will be helpful for law enforcement agencies in designing anti-fraud measures.

The rest of the article is structured as follows. Section 2 presents a literature review, Section 3 describes the data and methodology, Section 4 presents empirical findings, Section 5 discusses the results, and Section 6 concludes.

2. Literature Review

Financial fraud has been a global concern, where fraudsters deceive people to gain money or assets with various swindling methods. The United States Department of Justice [16] characterized fraud as deceiving people with the promise of goods, services, or other benefits, which did not exist or were never intended to be delivered. Financial frauds not only cause financial damages but also push people to depression and other psychological trauma [19,20]. Several large-scale financial frauds involving Ponzi schemes, pyramid investments, and hedge fund-related frauds have impacted the financial world so intensely that victims are still suffering from financial damage [21,22,23]. However, at the consumer level, telephone fraud, fictitious billing fraud, loan guarantee fraud, and refund fraud have spread sporadically over the last few decades, causing billions of dollars in financial loss to victims [24,25,26]. Despite several visible efforts by law enforcement agencies, incidents of financial fraud have been continuing to increase due to innovative swindling methods, changing target victims, changing the scope of the fraud, and using emerging financial and economic crises to design new types of financial fraud [24,27].

The ongoing COVID-19 pandemic has created a situation in which fraudsters are using new approaches to deceive consumers. The increasing use of online platforms during the pandemic has caused an increase in online transaction-related frauds [8,9,10]. Moreover, demand for several products related to the pandemic and safety measures had increased dramatically compared to their supply, causing significant friction and fraudulent activities such as non-delivery of goods, sending low-quality products, charging higher prices, etc. [7,10,11]. In addition, fraudsters distributed fake information to increase the stock price of companies that were engaged in pandemic-related goods and services [12]. Artificial increases in stock prices caused many people to suffer financial losses. Finally, fraudsters utilized several systems to swindle money that people received from the government’s stimulus packages [4]. Fraudsters stole sensitive financial information from victims pretending to be government officials distributing stimulus money and later used that information to swindle money from the victim’s account.

The incidence of financial fraud during the pandemic is not rare in Japan too. The changing socioeconomic situation has provided opportunities for fraudsters to devise new approaches to deceive people [17]. Particularly, online trading fraud and stimulus money distribution-related fraud became prolific in Japan [17,18]. Along with pandemic-related frauds, previously dominant financial frauds such as telephone fraud, loan guarantee fraud, and refund fraud were also evident in Japan [2,13].

While financial fraud is quite evident in the existing literature, there is a lack of study on victim profiles, particularly during the pandemic. A comparison between victims of financial fraud before and during the pandemic is also missing, but it is important to understand the mindset of fraudsters to target victims. However, a comprehensive study on victim profiles is always challenging because of underreporting of fraud cases, the fraudster’s innovative swindling mechanisms, and changes in target victims [3,24,28]. Although previous studies provided some common factors for fraud victimization, such as older age, being male, living in city areas, dissatisfaction with the current financial situation, and lack of conscientiousness [13,24,29,30,31,32], they fail to provide a reliable scenario for understanding specific financial fraud [2,33]. In many cases, victim diversity is observed in terms of gender, age, marital status, race, education, financial literacy, openness, etc. [28,34,35,36]. To fill the gap in the existing research, our objective is to compare the magnitude of financial frauds before and during the pandemic and to explore victim profiles during the pandemic. We hypothesize that risk factors for financial fraud during the pandemic differ at the aggregate and specific levels. Since fraudsters continuously change swindling methods, it is likely that they will not repeat similar mechanisms for all types of financial fraud.

Hypothesis 1 (H1).

The risk factors for financial fraud during the pandemic differ at the aggregate and specific levels.

Hypothesis 2 (H2).

The risk factors are different for specific types of financial fraud during the pandemic.

3. Data and Methodology

3.1. Data

A nationwide online panel survey conducted by the Hiroshima Institute of Health Economics Research, Hiroshima University, Japan, provides data for this study. The first wave of the survey was conducted in February 2020, just before the World Health Organization declared the COVID-19 situation a pandemic, while the second wave was conducted after a year in February 2021. Both waves surveyed victimization of financial frauds of different types along with other demographic, socioeconomic, and psychological backgrounds of respondents. The panel survey followed a random sampling procedure and ensured proper representation of data on important socio-demographic features. The Nikkei Research Company, a famous research company in Japan, used one of the largest nationwide databases to draw a sample in each wave. We had to exclude several observations due to missing values for socioeconomic variables. We checked the distribution of the excluded data, which is not significantly different from those used in the study. Thus, we believe that the exclusion of several data would materially affect the result of this study. The minimum age of respondents was set at 21 years, and the final sample size after merging the first and second waves was 4253. We used STATA statistical software to analyze data.

3.2. Variable Definitions

The dependent variable of this study is the victims of financial frauds. In the first wave, respondents were asked if they had experienced financial frauds, such as the “it’s me” fraud, fictitious billing fraud, loan guarantee fraud, or refund fraud, during the three years preceding the survey. However, in the second wave, respondents were asked if they had experienced any of those frauds during the last year, that is, during the pandemic. We considered the respondents as victims of financial frauds at the aggregate level if they experienced any kind of fraud during the concerned period. Moreover, we recognized the respondents as victims of financial frauds at a specific level if they experienced one of the specific frauds, such as the “it’s me” fraud, fictitious billing fraud, loan guarantee fraud, or refund fraud.

Independent variables in this study include the demographic, socioeconomic, and psychological characteristics of respondents. We included demographic characteristics as independent variables because previous studies found that factors like sex, age, marital status, and living with family are associated with victims of financial frauds [2,13,28,30,31,34]. Similarly, we included socioeconomic characteristics such as education, financial literacy, household income, household financial assets, residential status, and employment status as independent variables because previous studies found that people’s socioeconomic features were related to victims of financial frauds [2,13,28,34,36,37,38,39,40]. In particular, financial literacy was included in the study as an independent variable because Burton [35] and Ledbetter [41] found its association with financial frauds. We measured financial literacy following Lusardi and Mitchell’s [42] methodology, which is easy to understand and widely used in previous studies [43,44,45,46]. Finally, we included the psychological characteristics of respondents in independent variables because previous studies showed that factors like a myopic view of the future, level of financial satisfaction, anxiety about life in old age, careful buying habits, and trust in others had associations with financial frauds [2,33,47,48].

Table 1. Definition and measurement of variables.

Dependent Variable	Criteria
Financial fraud	Binary variable. 1 = participants have experienced financial frauds, such as the “it’s me” fraud, fictitious billing fraud, loan guarantee fraud, or refund fraud in the three years (last one year in the second wave) preceding the survey, 0 = otherwise.
Independent Variables
Male	Binary variable. 1 = male, 0 = female.
Age	Continuous variable. Age of respondents in years.
Married	Binary variable. 1 = married, 0 = otherwise.
Child	Binary variable. 1 = having a child/children, 0 = otherwise.
Living alone	Binary variable. 1 = respondents living alone, 0 = otherwise.
Living in rural areas	Binary variable. 1 = Living in rural areas (Not in Tokyo special wards or government-designated city areas), 0 = otherwise.
Education	Continuous variable. Years of education completed by respondents.
Financial literacy	Continuous variable. Financial literacy measures respondents’ ability to understand basic financial calculations, inflation, and risks of financial securities. The following questions were asked of respondents: 1. Suppose you have ¥100 in your savings account, the interest rate is 2% per year, and you never withdraw money or interest payments. After five years, how much will you have in this account? □ More than ¥102 □ Exactly ¥102 □ Less than ¥102 □ Do not know □ Refuse to answer 2. Assume that the interest rate on your savings account is 1% per year and inflation is 2% per year. After one year, how much will you be able to buy with the money in this account? □ More than today □ Exactly the same □ Less than today □ Do not know □ Refuse to answer 3. Indicate whether the following statement is true or false: “Buying a company’s stock usually provides a safer return than a stock mutual fund.” □ True □ False □ Do not know □ Refuse to answer
Employment status	Binary variable. 1 = currently employed, 0 = otherwise.
Household income	Continuous variable. Log of annual household income in yen.
Household assets	Continuous variable. Log of household financial assets in yen.
Myopic view of the future	Ordinal variable. Respondents’ perceptions about the future were measured by making them rate the following statement on a scale of 1 to 5: “As the future is uncertain, it is a waste of time thinking about it” (5 stands for “completely agree” and 1 for “completely disagree”).
Financial satisfaction	Ordinal variable. Respondents’ current level of financial satisfaction was measured by making them rate the following statement: “I am happy with my financial status” (5 stands for “completely agree” and 1 stands for “completely disagree”).
Anxiety	Ordinal variable. Respondents’ anxiety about life in old age was measured by making them rate the following statement: “I have anxieties about my life after I turn 65” (5 stands for “the highest level of anxiety” and 1 stands for “the lowest level of anxiety”).
Careful spending habit	Ordinal variable. Respondents’ carefulness in spending was measured by making them rate the following statement: “I think carefully before buying anything” (5 means “the respondent is the most careful about spending” and 1 means “the respondent is the least careful”).
Trust in people	Ordinal variable. Respondents’ trust in other people was measured by making them rate the following statement: “In general, most people are trustworthy” (5 stands for “completely agree” and 1 stands for “completely disagree”).

provides the definitions and measurements of all variables.

3.3. Descriptive Statistics

Table 2. Descriptive statistics.

Variable	Mean	Std. Dev.	Min.	Max.	Obs.
Financial fraud 2020	0.0543	0.2267	0	1	4253
Financial fraud 2021	0.0590	0.2357	0	1	4253
Male	0.6544	0.4756	0	1	4253
Age	50.3184	13.8258	21	86	4253
Married	0.6605	0.4736	0	1	4253
Child	0.5711	0.4950	0	1	4253
Living alone	0.2015	0.4012	0	1	4253
Living in rural areas	0.5815	0.4934	0	1	4253
Education	14.9697	2.1129	9	21	4253
Financial literacy	0.6524	0.3568	0	1	4253
Employment status	0.6381	0.4806	0	1	4253
Log of household income	15.4271	0.7598	13.12	16.86	4253
Log of household assets	15.8515	1.4297	14.04	18.64	4253
Myopic view of the future	2.6852	1.0174	1	5	4253
Financial satisfaction	2.7437	1.1153	1	5	4253
Future anxiety	3.7129	1.1380	1	5	4253
Careful buying habit	4.0165	1.0055	1	5	4253
Trust in people	2.8587	0.9577	1	5	4253

provides the descriptive statistics for the main variables of this study. The results show that 5.43% and 5.90% of respondents experienced financial fraud in 2020 and 2021, respectively. These numbers are quite significant in the context of Japan, as previous studies showed that most victims of financial frauds either did not report incidents of fraud or were very hesitant to share the incidents [2,15]. The demographic profile of the respondents shows that: (i) 65% of them are male, (ii) the respondents have an average age of 50.32 years (SD = 13.83 years), (iii) 66.05% of them are married, (iv) 57.11% of them have at least a child, and (v) 20.15% live alone. The socioeconomic status of the respondents shows that: (i) they have an average education of 14.97 years (SD = 2.11 years), (ii) they have an average financial literacy score of 0.65 out of 1 (SD = 0.36), (iii) 63.81% of them are currently employed, and (iv) they have a log of household income and assets of 15.43 (SD = 0.76) and 15.85 (SD = 1.43), respectively. Moreover, 58.15% of the respondents live outside government-designated city areas. The psychological orientation of the respondents reveals that: (i) they are moderately myopic about the future, (ii) they have a moderate level of satisfaction with their current financial status, (iii) they are relatively anxious about the future, and (iv) they have a careful buying behavior and a moderate level of trust in others.

Table 3. Comparison of financial frauds between pre-pandemic and pandemic periods.

Fraud Type	Pre-Pandemic (2020)	Pandemic (2021)	Difference (t Value)
Full Sample
Financial fraud at the aggregate level	0.0543	0.0590	0.0047 (0.94)
“It’s me” fraud	0.0127	0.0134	0.0007 (0.29)
Fictitious billing fraud	0.0315	0.0346	0.0031 (0.79)
Loan guarantee fraud	0.0122	0.0094	−0.0028 (1.26)
Refund fraud	0.0084	0.0113	0.0028 (1.32)
Younger subsample
Financial fraud at the aggregate level	0.0485	0.0555	0.0070 (1.57)
“It’s me” fraud	0.0095	0.0112	0.0017 (0.79)
Fictitious billing fraud	0.0294	0.0342	0.0048 (1.29)
Loan guarantee fraud	0.0126	0.0095	−0.0031 (−1.39)
Refund fraud	0.0065	0.0101	0.0036 (1.79) *
Older subsample
Financial fraud at the aggregate level	0.0845	0.0773	−0.0073 (−0.60)
“It’s me” fraud	0.0292	0.0248	−0.0044 (−0.73)
Fictitious billing fraud	0.0423	0.0364	−0.0058 (−0.65)
Loan guarantee fraud	0.0102	0.0088	−0.0015 (−0.33)
Refund fraud	0.0190	0.0175	−0.0015 (−0.23)
Male
Financial fraud at the aggregate level	0.0586	0.0661	0.0076 (1.39)
“It’s me” fraud	0.0162	0.0158	−0.0004 (−0.13)
Fictitious billing fraud	0.0327	0.0381	0.0054 (1.23)
Loan guarantee fraud	0.0186	0.0108	−0.0078 (−0.42)
Refund fraud	0.0093	0.0140	0.0047 (1.79) *
Female
Financial fraud at the aggregate level	0.0463	0.0456	−0.0007 (−0.10)
“It’s me” fraud	0.0061	0.0088	0.0027 (0.94)
Fictitious billing fraud	0.0293	0.0279	−0.0014 (−0.25)
Loan guarantee fraud	0.0129	0.0068	−0.0061 (−1.96) **
Refund fraud	0.0068	0.0061	−0.0007 (−0.23)

Note: Z values in parentheses. ** p < 0.05, * p < 0.1.

shows a detailed comparison of financial frauds between the pre-pandemic and pandemic periods for the full sample as well as four subsamples based on gender and age. The results of the full sample comparative analysis show that financial frauds have not significantly changed at the aggregate and specific levels during the pandemic compared to the pre-pandemic era. The results of the age-based subsamples show that financial frauds have not significantly changed at the aggregate level for younger and older subsamples. At specific levels, we observe no significant change except a considerable increase in refund frauds for the younger subsample. The gender-based subsample analysis shows that refund frauds increased significantly for males, while loan guarantee frauds decreased significantly for the female subsample.

Table 4. Victims of financial frauds at the aggregate level during the pandemic, stratified by gender and age.

	Male			Female			Male vs. Female
Type of Fraud	Younger (≤65 Years)	Older (>65 Years)	Younger vs. Older Male	Younger (≤65 Years)	Older (>65 Years)	Younger vs. Older Female
Financial fraud at the aggregate level	0.0624	0.0807	−0.0183 (−1.57)	0.0443	0.0603	−0.0160 (−0.79)	0.0205 (2.70) ***
“It’s me” fraud	0.0127	0.0281	−0.0154 (−2.63) ***	0.0089	0.0086	−0.0002 (−0.03)	0.0070 (1.88) *
Fictitious billing fraud	0.0375	0.0404	−0.0029 (−0.32)	0.0288	0.0172	0.0116 -0.73	0.0102 (1.73) *
Loan guarantee fraud	0.0118	0.0070	0.0047 −0.98	0.0118	0.0070	−0.0047 (−0.98)	0.0040 −1.28
Refund fraud	0.0131	0.0175	−0.0044 (−0.80)	0.0052	0.0172	−0.0121 (−1.60)	0.0079 (2.32) **

Note: t values in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

shows victims of financial frauds during the pandemic stratified by gender and age. The results show that older males became victims of the “it’s me” fraud significantly more than their younger counterparts. A comparison between male and female subsamples shows that males became victims of all types of frauds significantly more than females except loan guarantee fraud. The results show that the gender and age of victims have important consequences for the study of financial frauds during the pandemic.

3.4. Methods

We used the following equations to estimate the association between victims of financial frauds and demographic, socioeconomic, and psychological variables:

Y1i = f (Xi, ei)

(1)

Y2i = f (Xi, ei),

(2)

where Y1 is the number of victims of financial frauds during the pandemic and Y2 is the number of new victims of financial frauds during the pandemic who had not experienced financial frauds before. X is the vector of demographic, socioeconomic, and psychological variables during the pandemic, which includes gender, age, marital status, having a child, living status, living in a rural area, education, financial literacy, employment status, household income, the value of household assets, a myopic view of the future, satisfaction about current financial status, anxiety about the future, careful buying habits, and trust of other people. As the dependent variables are binary, logit regression models were used to estimate Equations (3) and (4).

Additionally, we measured VIF to test multicollinearity among explanatory variables (results are not included to save space but are available upon request). The results show that the VIFs of explanatory variables are well below 10. Thus, regression results are unlikely to be affected by the multicollinearity problem.

The full model specifications for Equations (1) and (2), which use four types of financial frauds, are as follows:

logit {P(Victims of financial fraudi = 1|Xi)} = β0 + β1 malei + β2 agei + β3 spousei + β4 childreni + β5 living alonei + β6 living in rurali + β7 educationi + β8 employedi + β9 log (household incomei) + β10 log (household assetsi) + β11 financial literacy + β12 myopic view of the future + β13 future anxiety + β14 financial satisfactioni + β15 careful buying habit +β16 trust

(3)

logit {P(New victims of financial fraudi = 1|Xi)} = β0 + β1 malei + β2 agei + β3 spousei + β4 childreni + β5 living alonei + β6 living in rurali + β7 educationi + β8 employedi + β9 log (household incomei) + β10 log (household assetsi) + β11 financial literacy + β12 myopic view of the future + β13 future anxiety + β14 financial satisfactioni + β15 careful buying habit +β16 trust

(4)

4. Empirical Findings

Table 5. Estimation results of victims of financial frauds during the pandemic.

Variable	Financial Fraud (Aggregate Level)	“It’s Me” Fraud	Fictitious Billing Fraud	Loan Guarantee Fraud	Refund Fraud
Male	0.4618 (2.65) ***	0.7646 (2.06) **	0.3621 (1.61)	0.7107 (1.68) *	0.9381 (2.20) **
Age	−0.0062 (−0.98)	−0.0161 (−1.26)	0.0000 (0.01)	−0.0392 (−2.57) **	−0.0037 (−0.25)
Married	−0.1251 (−0.62)	0.0140 (0.03)	−0.0948 (−0.36)	−0.8171 (−1.73) *	0.1226 (0.26)
Child	0.1016 (0.58)	0.1598 (0.44)	0.0529 (0.24)	0.3780 (0.85)	−0.0658 (−0.17)
Living alone	−0.7230 (−3.04) ***	−0.4964 (−1.02)	−0.5843 (−1.92) *	−0.7991 (−1.49)	−0.2991 (−0.56)
Living in rural areas	−0.0508 (−0.38)	0.0773 (0.28)	0.0068 (0.04)	−0.2491 (−0.77)	−0.1910 (−0.64)
Education	0.0030 (0.09)	−0.0349 (−0.51)	−0.0050 (−0.12)	0.0417 (0.51)	−0.0836 (−1.11)
Employment status	0.2760 (1.60)	0.2906 (0.81)	0.1922 (0.87)	0.1121 (0.25)	0.0934 (0.24)
Log of household income	−0.2538 (−2.30) **	−0.3438 (−1.55)	−0.0989 (−0.67)	0.2439 (0.79)	−0.0781 (−0.30)
Log of household assets	0.2502 (4.19) ***	0.3090 (2.52) **	0.2453 (3.20) ***	0.1682 (1.14)	0.4240 (3.16) ***
Financial literacy	−0.6889 (−3.47) ***	−0.8310 (−2.09) **	−0.5617 (−2.19) **	−0.3188 (−0.66)	−0.5068 (−1.13)
Myopic view of the future	0.1301 (1.94) *	0.3078 (2.25) **	0.0929 (1.09)	0.2591 (1.61)	0.1274 (0.87)
Future anxiety	0.0794 (1.15)	0.0289 (0.21)	0.1849 (2.01) **	−0.1257 (−0.82)	0.1241 (0.81)
Financial satisfaction	−0.0074 (−0.10)	0.2332 (1.54)	−0.1109 (−1.18)	0.0774 (0.45)	−0.0188 (−0.11)
Careful buying habit	−0.2177 (−3.34) ***	−0.2014 (−1.55)	−0.1933 (−2.28) **	−0.1037 (−0.66)	−0.0620 (−0.42)
Trust on people	0.0413 (0.56)	0.1494 (0.96)	−0.0093 (−0.10)	−0.0193 (−0.11)	0.0263 (0.16)
Constant	−2.3741 (−1.44)	−4.2484 (−1.28)	−5.3986 (−2.45) **	−9.7307 (−2.20) **	−9.5749 (−2.43) **
Observations	4253	4253	4253	4253	4253
Log likelihood	−921.16	−284.91	−623.58	−213.39	−252.50
LR Chi2	65.19 ***	35.02 ***	30.99 **	26.16 *	20.94
Pseudo R2	0.0342	0.0579	0.0242	0.0577	0.0398

Note: Z values in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

shows the regression results of victims of financial frauds at the aggregate and specific levels during the pandemic. The results show that being male, living with family, having a low household income, having a higher value of household assets, having low financial literacy, having a myopic view of the future, and having careless buying habits were associated with the probability of being victims of financial frauds at the aggregate level. At specific levels, we found that being male, having a higher value of household assets, low financial literacy, and a myopic view of the future were associated with the “it’s me” fraud. At the same time, living with family, having a higher amount of household assets, low financial literacy, future anxiety, and careless buying habits were associated with fictitious billing fraud. Meanwhile, being male, being young, and being unmarried were associated with loan guarantee fraud. Moreover, being male and having a higher amount of household assets were associated with the refund fraud.

As victims of financial frauds during the pandemic include both new victims (those people who had never been a fraud victim before the pandemic) and repeat victims (those people who had been a fraud-victim before the pandemic), we also estimated regression coefficients for new victims of financial frauds. The estimation results in

Table 6. Estimation results of new victims of financial frauds during the pandemic.

Variable	Financial Fraud (Aggregate Level)	“It’s Me” Fraud	Fictitious Billing Fraud	Loan Guarantee Fraud	Refund Fraud
Male	0.3166 (1.55)	0.6238 (1.48)	0.2527 (1.01)	0.9720 (1.87) *	0.8086 (1.85) *
Age	−0.0033 (−0.44)	−0.0387 (−2.47) **	0.0003 (0.04)	−0.0411 (−2.33) **	−0.0100 (−0.66)
Married	−0.0951 (−0.40)	0.3058 (0.61)	−0.1224 (−0.42)	−0.8241 (−1.53)	0.0100 (0.02)
Child	−0.1605 (−0.79)	−0.1173 (−0.28)	−0.0446 (−0.18)	0.5126 (1.00)	−0.1584 (−0.39)
Living alone	−0.7179 (−2.63) ***	−0.2281 (−0.40)	−0.4850 (−1.46)	−1.0831 (−1.72) *	−0.3933 (−0.70)
Living in rural areas	0.0061 (0.04)	0.2336 (0.68)	−0.0406 (−0.21)	−0.2689 (−0.72)	−0.2448 (−0.77)
Education	−0.0006 (−0.01)	−0.1257 (−1.52)	0.0125 (0.25)	0.0218 (0.23)	−0.0918 (−1.14)
Employment status	0.4903 (2.34) **	0.4199 (0.92)	0.4822 (1.85) *	0.4026 (0.76)	0.1001 (0.24)
Log of household income	−0.2496 (−1.89) *	−0.0771 (−0.26)	−0.0695 (−0.41)	−0.1260 (−0.39)	0.1050 (0.36)
Log of household assets	0.1663 (2.35) **	0.3356 (2.21) **	0.1856 (2.14) **	0.2090 (1.23)	0.3912 (2.71) **
Financial literacy	−0.3135 (−1.31)	−0.5739 (−1.21)	−0.3185 (−1.08)	−0.3338 (−0.60)	−0.3664 (−0.76)
Myopic view of the future	0.0912 (1.15)	0.2583 (1.57)	0.0426 (0.44)	0.1057 (0.57)	0.1801 (1.15)
Future anxiety	0.1201 (1.45)	0.0553 (0.34)	0.1870 (1.81) *	−0.1566 (−0.89)	0.1002 (0.63)
Financial satisfaction	0.0547 (0.63)	0.2939 (1.64)	−0.0813 (−0.77)	−0.0414 (−0.21)	0.0567 (0.33)
Careful buying habit	−0.2240 (2.88) ***	−0.2789 (−1.80) *	−0.2601 (−2.76) ***	−0.1947 (−1.10)	−0.1168 (−0.76)
Trust on people	0.0298 (0.34)	0.0139 (0.08)	0.0327 (0.31)	−0.0839 (−0.41)	−0.0604 (−0.35)
Constant	−1.9237 (−0.98)	−6.5346 (−1.53)	−5.4492 (2.16) **	−3.5421 (−0.77)	−11.2110 (−2.59) **
Observation	4253	4253	4253	4253	4253
Log likelihood	−702.19	−206.21	−509.61	−168.09	−225.82
LR Chi2	36.08 ***	31.19 **	23.68 *	20.86	19.83
Pseudo R2	0.0250	0.0703	0.0227	0.0584	0.0421

Note: Z values in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

show that respondents who live with their family, are currently employed, have a low household income, have a higher value of household assets, and have careless buying habits are associated with the probability of being new victims of financial frauds at the aggregate level. At specific levels, we found that respondents who were young, had a higher value of household assets, and had careless buying habits become new victims of the “it’s me” fraud. Respondents who were currently employed, had a higher value of household assets, were anxious about the future, and had careless buying habits became new victims of the fictitious billing fraud. Respondents who were male, younger in age, and lived with families became new victims of the loan guarantee fraud. Respondents who were male and had a higher value of household assets became new victims of the refund fraud.

Finally, we re-estimated regression models using three interaction variables based on demographic variables like being male, age, and living alone because these variables are found to be significantly associated with financial frauds at the aggregate and specific levels.

Table 7. Estimation results of new victims of financial frauds during the pandemic using interaction variables.

Variable	Financial Fraud (Aggregate Level)	It’s Me Fraud	Fictitious Billing Fraud	Loan Guarantee Fraud	Refund Fraud
Male	−0.6969 (−0.99)	1.1435 (0.81)	−2.1209 (−2.39) **	−0.2437 (−0.14)	3.0074 (1.98) *
Age	−0.0126 (−0.96)	−0.0181 (−0.69)	−0.0335 (−1.91) *	−0.0573 (−1.43)	0.0198 (0.73)
Married	−0.1171 (−0.49)	0.3066 (0.61)	−0.1701 (−0.57)	−0.8409 (−1.55)	0.0953 (0.20)
Child	−0.1498 (−0.74)	−0.1106 (−0.27)	−0.0232 (−0.09)	0.5128 (1.00)	−0.1685 (−0.42)
Living alone	−0.9704 (−1.10)	−13.1160 (−0.01)	−0.6965 (−0.71)	−13.5435 (−0.01)	−3.0250 (−1.26)
Living in rural areas	0.0024 (0.02)	0.2565 (0.75)	−0.0550 (−0.28)	−0.2631 (−0.71)	−0.2413 (−0.76)
Education	−0.0026 (−0.06)	−0.1211 (−1.46)	0.0054 (0.11)	0.0222 (0.24)	−0.0818 (−1.01)
Employment status	0.5591 (2.66) ***	0.5873 (1.26)	0.5453 (2.09) **	0.4826 (0.91)	0.1162 (0.27)
Log of household income	−0.2498 (−1.91) *	−0.1156 (−0.40)	−0.0507 (−0.3)	−0.1224 (−0.38)	0.0848 (0.29)
Log of household assets	0.1652 (2.33) **	0.3242 (2.13) **	0.1931 (2.22) **	0.2044 (1.21)	0.3765 (2.63) ***
Financial literacy	−0.2927 (−1.22)	−0.5860 (−1.24)	−0.2626 (−0.88)	−0.3177 (−0.57)	−0.3893 (−0.82)
Myopic view of the future	0.0940 (1.18)	0.2472 (1.50)	0.0557 (0.57)	0.1046 (0.56)	0.1727 (1.10)
Future anxiety	0.1188 (1.42)	0.0588 (0.36)	0.1850 (1.78) *	−0.1581 (−0.89)	0.1033 (0.64)
Financial satisfaction	0.0550 (0.64)	0.2999 (1.68) *	−0.0842 (−0.79)	−0.0402 (−0.20)	0.0546 (0.31)
Careful buying habit	−0.2239 (−2.88) ***	−0.2844 (−1.82) *	−0.2571 (−2.73) ***	−0.1972 (−1.11)	−0.1208 (−0.78)
Trust on people	0.0276 (0.31)	0.0053 (0.03)	0.0344 (0.32)	−0.0863 (−0.42)	−0.0678 (−0.39)
Malexage	0.0178 (1.23)	−0.0199 (−0.66)	0.0494 (2.60) ***	0.0239 (0.56)	−0.0459 (−1.59)
Malexlivealone	1.2391 (2.00) **	14.1828 (0.01)	0.8615 (1.22)	13.4421 (0.01)	0.5151 (0.42)
Agexlivealone	−0.0147 (−0.81)	−0.0208 (−0.53)	−0.0085 (−0.38)	−0.0156 (−0.34)	0.0437 (1.19)
Constant	−1.4112 (−0.71)	−6.5576 (−1.54)	−4.3535 (−1.70) *	−2.7817 (−0.59)	−12.1730 (−2.74) ***
Observation	4253	4253	4253	4253	4253
Log likelihood	699.30	−203.44	−505.41	−166.85	−223.80
LR Chi2	41.86 ***	36.73 ***	32.07 **	23.34	23.86
Pseudo R2	0.0291	0.0828	0.0308	0.0654	0.0506

Note: Z values in parentheses. *** p < 0.01, ** p < 0.05, * p < 0.1.

shows the estimation results with three interaction variables, such as “malexage”, “malexlivealone”, and “agexlivealone”. The results show that the variable “malexlivealone” is positively associated with new victims of financial frauds at the aggregate level, indicating that respondents who are male and live alone are more likely to be new victims of financial frauds. The other two interaction variables do not have significant associations with financial frauds at the aggregate level. At specific levels, the interaction between being male and age has a significantly positive association with new victims of the fictitious billing fraud, indicating that respondents who are male and older in age have higher chances of being new victims of the fictitious billing fraud. Other interaction variables are not significantly associated with any specific financial fraud.

In summary, the results show that the factors associated with the probability of being victims of financial frauds at the aggregate and specific levels differ to some extent. Moreover, differences are found among specific types of financial frauds, that is, the “it’s me” fraud, fictitious billing fraud, loan guarantee fraud, and refund fraud. The differences are also articulated for all victims and new victims during the pandemic.

5. Discussion

The disruptions in social and economic lives caused by the pandemic created an opportunistic environment for financial fraudsters to swindle money from people through innovative mechanisms using the necessities during the pandemic. In particular, the proliferation of online trading due to restrictions on movement and work-from-home, changes in the relative demand for goods, changes in socioeconomic conditions, and government stimulus programs appear to provide ground for using new swindling mechanisms [4,5,8,9,11,17,18]. In this study, we sought to explain the socioeconomic background of victims of financial frauds during the pandemic with an emphasis on new victims. However, the underreporting of fraud cases has always been a concern for studies on financial frauds. The reported cases of frauds are traditionally compared to the tip of the iceberg. Due to the lack of knowledge on fraud reporting and some social reasons, people become demotivated to report fraud cases [13,15,28,39]. Even then, fraud cases that were reported during the pandemic will help to understand the risk factors associated with victims of financial frauds.

We found that the instances of the refund fraud have increased for men, while those of the loan guarantee fraud have decreased significantly for women during the pandemic. We observed no significant change for other fraud types. The evidence of financial frauds during the pandemic should be interpreted pragmatically. The proliferation of financial frauds was observed in the context of COVID-19-related products and services [49,50,51]. It appears that fraudsters focused less on traditional channels of financial frauds, which could be a reason for a reduction in the number of loan guarantee fraud cases for women. Moreover, disruptions in economic activities could also have reduced the demand for loan and credit cards. The increase in the refund fraud cases for men, particularly among the younger subsample, during the pandemic is concerning because previous studies mostly found such cases among older people due to a lack of technological knowledge, lower cognitive ability, higher emotion, and higher familial engagement [2,12,13,15,18,19,30,32].

We estimated the association between demographic, socioeconomic, and psychological factors associated with the probability of being victims of financial frauds at the aggregate and specific levels. At the aggregate level, we found that being male, living with family, having a lower income, having a higher value of financial assets, having lower financial literacy, having a myopic view of the future, and having careless buying habits increased the odds of being victims of financial frauds. However, there are similarities and dissimilarities in factors associated with financial frauds at the aggregate and specific levels. For example, being male is also associated with the victims of the “it’s me”, loan guarantee, and refund frauds; younger age and being married are associated with the loan guarantee fraud; living with family is associated with the fictitious billing fraud; the value of household assets is associated with the “it’s me”, fictitious billing, and refund frauds; financial literacy is associated with the “it’s me” and fictitious billing frauds; a myopic view of the future is associated with the “it’s me” fraud; and anxiety about the future and careless buying habits are associated with the fictitious billing fraud. Previous studies reported the higher probability of men being victims of certain financial frauds and attributed it to their greater tendency to take risks and greater familial responsibility than women [2,13,36]. However, compared to the conventional belief that older people have a higher probability of being victims, we found that younger and unmarried people have a higher likelihood of becoming victims of loan guarantee frauds. Meanwhile, Kadoya et al. [2] found evidence that older people being victims of financial frauds was more related to the “it’s me” frauds and refund frauds than loan guarantee frauds. The association between living with family and the victims of fictitious billing frauds has also been documented earlier in Japan [2] and attributed to higher online transactions due to familial reasons.

Regarding socioeconomic factors, the higher probability of low-income people being the victims of financial frauds is supported by previous studies [2,39]. These studies attributed this tendency to low-income people possibly venturing into risky avenues for earning money. At the same time, wealthy people are usually targets of fraudsters because of their higher asset values, which is evident in previous studies [2,13]. Financial literacy, widely recognized as a rational-decision making tool [46,52,53], reduces the probability of being victims of the “it’s me” and fictitious billing frauds, as it enables people to have better financial judgment [2]. Finally, psychological factors such as a myopic view of the future and uncareful buying habits, which have associations with the “it’s me” frauds and fictitious billing frauds, respectively, show that people’s impulsive behavior, lack of judgment, and susceptibility to persuasion drive them to be victims of frauds [41,54].

Victims of financial frauds during the pandemic include both new and repeat victims. To identify new victims during the pandemic, we removed repeat victims and re-estimated the regression models. The results show unique characteristics of new victims of financial frauds during the pandemic. Compared to all victims of financial frauds, being male is not associated with the victims of “it’s me” frauds, but associated with loan guarantee and refund frauds; younger age is associated with “it’s me” frauds along with loan guarantee frauds; being married is not associated with loan guarantee frauds; living with family is not associated with billing frauds, but associated with loan guarantee frauds; employment status is associated with billing frauds; financial literacy is not associated with “it’s me” and fictitious billing frauds; a myopic view of the future is not associated with “it’s me” fraud; future anxiety is associated with fictitious billing frauds; and careless buying habits is associated with “it’s me” frauds along with fictitious billing frauds. We found some important insights on the covariates of financial frauds for new victims. For example, against the conventional belief that older people experience “it’s me” frauds [2], our study revealed that younger people became victims of “it’s me” frauds during the pandemic. Moreover, victims of fictitious billing frauds were currently employed, had a higher value of household assets, and had uncareful buying habits but were anxious about the future. When interaction terms were added to the estimation, results showed that older males, females, and younger people, along with previously reported employed, wealthy, and anxious people, became victims of fictitious billing frauds. These findings indicate that the victims were financially well-off but went for panic buying and did not carefully verify the authenticity of the suppliers. Finally, unlike previous situations, our study showed that financial literacy did not play any role in predicting financial fraud, which means that new victims did not rationally make their financial and consumer decisions during the pandemic. Our findings are consistent with several studies reporting anxiety, panic buying, and irrational hoarding during the pandemic [55,56,57].

Our study has some limitations, which should be considered while interpreting the results. First, recognition and reporting bias is a common phenomenon in the study of financial fraud. People often do not recognize when they experience a fraud case or report it appropriately for fear of social embarrassment. As a result, the number of fraud cases for individual fraud types is rather low, which could impact the statistical significance of the logit model. However, given the scenario of low fraud reporting, our study provides important evidence for the authorities even with lower statistical power. Second, we could not identify fraud cases specific to COVID−19-related products and services due to the unavailability of data. Future studies should investigate more specific types of frauds that people experienced during the pandemic. Moreover, the experience of frauds should be studied together with sociopsychological phenomena, such as panic, trauma, and social isolation. Third, Ai and Norton [58] argued that the impact of interaction terms in a logit model could not be perfectly evaluated by simply observing its sign, magnitude, or statistical significance. In our discussion of results, we presented a possible outcome of the interaction term and requested readers to consider possible misrepresentation as claimed by Ai and Norton [58]. Future research on financial fraud should consider all these limitations and focus more on how people’s economic motivations are affected by the incidents of financial fraud.

6. Conclusions

The changing social and economic environments and consequent changes in the financial and psychological conditions of people have created a suitable environment for financial fraud scammers. The increasing trend of COVID-19-related frauds requires understanding the probable risk factors and identifying people at risk. This study compares incidences of “it’s me” frauds, fictitious billing frauds, loan guarantee frauds, and refund frauds between the pre-pandemic and pandemic periods and identifies the covariates of financial frauds in Japan. The results show that refund frauds for men have increased, while loan guarantee frauds for women have decreased during the pandemic. The regression results show that attributes like being male, younger in age, living with family, employment status, household income, the value of household assets, financial literacy, a myopic view of the future, and careful buying habits are associated with the probability of being victims of financial frauds during the pandemic. In support of our hypothesis, this study also reports differences in covariates of financial frauds at the aggregate and specific levels, implying that fraudsters target a specific group of people for a particular type of financial fraud.

This study contributes to the understanding of financial fraud during the ongoing COVID-19 pandemic. Our results show that fraudsters target people with specific socioeconomic backgrounds for a particular financial fraud. Risk factors are not quite common among the types of financial fraud. We rather found significant differences among victims of fianncial frauds of various types. Furthermore, we found that fraudsters frequently changed the target group, making the swindling technique difficult to follow. We provide evidence that fraudsters devise swindling methods based on the worries and discomforts of victims and that sometimes fraudsters utilize changing socioeconomic situations to devise a new type of financial fraud. The results of this study imply that various sectors of people in society became affected differently due to the COVID-19 pandemic, which made them susceptible to different types of financial frauds. For example, employment status is associated with fictitious billing fraud but not with others, and future anxiety is associated with fictitious billing fraud but not with others, financial satisfaction is associated with telephone fraud but not with others, etc. Moreover, the traditional belief that being male and older in age are universal risk factors for financial fraud has not been validated in this study. Thus, a one-size-fits-all measure to combat financial frauds will not be effective. Policymakers and law enforcement agencies should educate people on individual fraud cases and possible swindling methods that fraudsters could apply. For example, the finding that older people are more susceptible to being victims of refund fraud requires concerned authorities in charge of distributing refunds, particularly COVID-19-related payments, to provide safeguards against older people. Financial institutions should also provide anti-fraud guidelines at the time of opening an account or at the time of conducting online transactions so that people can identify fraudulent activities. Finally, we emphasize people’s psychological soundness and ability to behave rationally; otherwise, they could be easy targets of fraudsters.