Product Counterfeits in the Online Environment

Abstract

The growing pace of e-commerce has facilitated the sale and distribution of counterfeit products. One reason may be that consumers cannot fully validate goods for sale online, thus creating tremendous opportunities for fraud. Despite the growth of online product counterfeiting specifically, little research has examined this crime which limits our basic understanding of the problem and victim reporting. Drawing on 2009 and 2010 complainant data from the Internet Crime Complaint Center, we examine the characteristics, costs, and reporting of online auction and non-auction product counterfeiting incidents. In light of the limitations of this study, we discuss the contribution of our findings for advancing theory and future research.

Keywords

product counterfeiting cybercrime criminal victimization

Introduction

The proliferation of computers and Internet-connected devices has revolutionized the way that individuals purchase goods and services (Newman & Clarke, 2003). Consumers can readily acquire a range of products through websites and auction services from businesses with brick and mortar locations off-line as well as individuals who sell direct to other consumers. This innovation, however, has drastically increased the ease with which individuals can engage in various forms of fraud, particularly product counterfeiting where individuals attempt to pass a manufactured copy (i.e., fake) of an original retail item in an attempt to deceive consumers (Dolan, 2004; Newman & Clarke, 2003; Wall & Large, 2010).

In general, product counterfeiting is a large and growing problem. Although estimates of its extent and costs are imprecise and challenged on methodological grounds (United States General Accountability Office [USGAO], 2010a), reports suggest that counterfeit products result in hundreds of billions of dollars in losses (see International Anti-Counterfeiting Coalition [IACC], 2005) and make up at least 5% of world trade (ICC Counterfeiting Intelligence Bureau, 1997). Although luxury items (e.g., branded apparel, jewelry, purses, sunglasses, etc.) are commonly counterfeited, virtually any product is vulnerable to counterfeiting including prescription medicines, automotive parts, electronics, sporting equipment, toys, batteries, hygiene products, alcohol, and cigarettes (Albanese, 2011; Nasheri, 2005; Phillips, 2005). Product counterfeiting poses serious consequences for many—consumers face health and safety risks (e.g., death or injury resulting from substandard goods), industries lose billions in revenue (Bate, 2008; MEMA, 2009) and suffer damages to brand reputation, and governments lose tax dollars (Freeman, Sidhu, & Montoya, 2007; Thompson, 2004) and must spend millions on intellectual property enforcement (USGAO, 2010a, 2010b). Finally, there is increasing concern over emerging evidence that international organized criminals and terrorist groups generate revenue from product counterfeiting (Albanese, 2011; Chow, 2003; Kelly, 2005; Liang, 2006; OECD, 2007).

In the online environment, e-commerce sites, such as eBay and half.com, enable offenders to use deceptive advertising practices like false images and persuasive language to sell counterfeit goods to unknowing customers (Dolan, 2004; Newman & Clarke, 2003; Wall & Large, 2010). In addition, these virtual environments do not allow consumers to inspect and fully validate the authenticity of items before purchasing them. Instead, buyers must depend upon a seller’s reputation and/or customer reviews to determine whether the seller is reliable and that an item will be delivered and genuine (Chua, Wareham, & Robey, 2007; Gregg & Scott, 2006).

Indeed, the ease of online advertising coupled with the low perceived risk for cybercrime offending has led to a significant increase in online product counterfeiting over the last decade (Wall & Large, 2010). For example, the Organization for Economic Cooperation and Development (OECD, 2007) suggested that billions of dollars are lost from the distribution of counterfeit products over the Internet. In some cases, however, an online seller may never actually deliver the product after receiving payment from the buyer. In fact, the Internet Crime Complaint Center (2011) reports nondelivery of goods as one of the most common complaints from victims over the last decade. Finally, the Federal Trade Commission (2011) reported that there were over 56,000 consumer complaints of Internet auction fraud in 2010 alone.

Despite the growth and varied harms of auction fraud and online product counterfeiting specifically, few researchers have explored the demographic composition of such victims (see Dolan, 2004; Newman & Clarke, 2003; Wall & Large, 2010). This stems in large part from the underreporting of this form of cybercrime to law enforcement agencies (Dolan, 2004). Many researchers have noted that victims may not know what agency has the proper jurisdiction to investigate complaints of fraud and related offenses, thereby reducing the likelihood of reporting (Dolan, 2004; Holt, 2003; Mason & Benson, 1996). Alternatively, an individual may feel embarrassed that they were defrauded or scammed, or even believe that small financial losses make the incident too unimportant to report to law enforcement (Dolan, 2004; Schoepfer & Piquero, 2009). Finally, some might think that local law enforcement is unable to investigate such incidents thereby making reporting moot (Holt, 2003; Nerenberg, 2000; Shichor, Sechrest, & Doocy, 2000).

Because of these factors there is a lack of substantive knowledge about the victims of online product counterfeiting and the factors affecting the reporting of this crime to law enforcement and other entities. To directly address this gap, this exploratory study uses complainant data reported to the Internet Crime Complaint Center (IC3) in 2009 and 2010 to identify the characteristics of suspects, victims, and transactions as well as the correlates of victim reporting. We discuss the shortcomings of these data but, especially in light of limited available data, also how this study advances theory and future research.

The Impact of the Internet on Product Counterfeiting

The emergence of e-commerce has directly affected the capacity of consumers to acquire new goods directly from a variety of retailers and secondhand products from other consumers. In turn, offenders have seized upon this opportunity to sell a variety of counterfeit products through online outlets. Product counterfeiters produce a variety of goods, with luxury items like clothing, sneakers, and jewelry a common target (Raustiala & Sprigman, 2005; Wall & Large, 2010). It is estimated that luxury goods account for no more than 5–10% of counterfeited items (see Phillips, 2005); however, these types of items may be particularly attractive to both online counterfeiters and consumers.

The Internet is a critical tool for the distribution of counterfeit products, including imperfect, lower-quality goods that resemble the original retail item (Lai & Zaichkowsky, 1999; Wall & Large, 2010). Counterfeit products are different from replicas or reproductions that are marketed as look-alike products in that they are presented as the authentic goods using language to entice the customer (Wall & Large, 2010).¹ Using online retail outlets, fraudsters can readily use images, logos, branding, and deceptive marketing strategies to make counterfeit goods appear legitimate to potential buyers. The prices for counterfeit goods are also low as to entice consumers who want to stay current with fashion trends and give the impression of higher social status by owning what look like luxury brand goods (Wall & Large, 2010).

Online environments are attractive to product counterfeiters because the opportunities for consumers to properly inspect an item are severely limited, making it difficult to detect fraudulent items before making a purchase. For example, counterfeiters commonly use spam e-mail to send advertisements for clothing or pharmaceuticals to unsuspecting victims (Balsmeier, Bergiel, & Viosca, 2004; Wall, 2004; Wood, 2004). E-mail is a key resource for counterfeiters because it is a cheap, efficient, and easily anonymized form of communication that can be exploited to contact larger populations with ease (Gartner, Inc., 2003). The senders use images including legitimate brand logos to create advertisements for the desired product along with sentiments that speak to the value and low cost of their merchandise (Balsmeier et al., 2004; Wall & Large, 2010). In turn, consumers must self-evaluate the authenticity of the advertisement in order to determine the legitimacy of a product and whether or not to purchase it.

Similarly, counterfeit goods are often sold through online auction sites like eBay where consumers must consider indicators of trust and reputation that can be artificially inflated or manipulated (Dolan, 2004). For example, an eBay seller profile can be hacked and hijacked by an individual in order to present the seller as a reputable member with many previous satisfied customers (Chua et al., 2007; Gregg & Scott, 2006). It is also difficult for consumers at auction sites to report their losses and receive compensation after they have determined that the product they purchased is counterfeit. eBay, for instance, does not offer direct compensation to customers, though they will log complaints against a seller. Sellers, however, can also create accounts using false names or addresses, making it difficult to immediately identify the actors responsible for the fraudulent exchange (Gregg & Scott, 2006).

One of the few studies to examine the correlates of auction fraud used a survey of individuals who reported auction fraud victimization to the National White Collar Crime Center (NW3C; Dolan, 2004). This study found that the average victim was a 30- to 40-year-old male with a bachelor’s degree making between $50,000 and $60,000 a year. The most common type of fraud experienced was the nondelivery of an item they paid for in full, followed by receiving a counterfeit good. Individuals who were defrauded participated in auctions at least once a month or week, indicating that they were regular auction users. The average dollar loss reported by victims ranged from $100 to $299.99, and nearly three fourths of individuals reported being victimized through eBay. In addition, respondents were very dissatisfied with both law enforcement and auction houses, indicating that very little had been done to resolve their complaint. At the same time, the respondents did not stop participating in auctions, as 77% engaged in auctions after their victimization.

Dolan’s (2004) study used a convenience sample of 49 complainants who contacted the NWC3 in 2003 regarding auction fraud. Though this study provides context for the phenomenon of auction fraud, it gives relatively little detail on the prevalence or incidence of fraud in the larger population. Additionally, it is largely descriptive in nature and gives no context for the factors influencing fraud reporting. Thus, the present study greatly expands upon Dolan’s work by utilizing a multiyear sample of all auction and non-auction fraud complainants who contacted the NWC3, rather than a subsample of this group. Furthermore, we employ regression techniques in order to address the correlates of reporting behavior, which is underexamined in the larger literature (e.g., Schafer & Olsen, 1998).

Predictors of Reporting Product Counterfeiting Victimization

The lack of empirical research on product counterfeiting limits our understanding of the factors that may affect victim reporting to police and regulatory agencies. There are, however, some similarities between product counterfeiting and electronic fraud victimization. There are myriad forms of fraud perpetrated through electronic communications like e-mail and online environments generally in order to obtain funds from unsuspecting victims through misrepresentation, enticing language, or fear (Holt & Graves, 2007; Holtfreter, Reisig, & Blomberg, 2006; James, 2005; Wall, 2004). Research on the latter, which incorporates instances of goods and services that are grossly misrepresented to victims (e.g., Holtfreter et al., 2006; Schoepfer & Piquero, 2009), may therefore provide insight into the nature of victim reporting for the former.

To that end, there appear to be few consistent correlates for reporting fraud to law enforcement or to other entities. In fact, some researchers have considered why fraud victims do not report their experiences (Mason & Benson, 1996; Van Wyk & Mason, 2001; Vaughan & Carlo, 1976). Many victims feel that they will not receive restitution for their losses and therefore find reporting to be an unnecessary burden (Vaughan & Carlo, 1976). In addition, there is relatively high dissatisfaction with law enforcement processes and practices among victims of fraud (Shichor et al., 2000). This may account for the low reporting rate noted across multiple studies of fraud victimization (Bass & Hoeffler, 1992; Blum, 1972; Holtfreter et al., 2006; Mason & Benson, 1996; Rebovich & Layne, 2000; Schoepfer & Piquero, 2009).

Several situational factors might influence whether or not product counterfeiting victims report. For example, a victim’s economic situation or the amount lost may affect their willingness to report their experience. Though the literature is mixed, some evidence suggests that victims from middle- and high-income groups are more likely to report fraud victimization (Baum, 2006; Blum, 1972). Because product counterfeiters often target advertisements at individuals who desire products used by those in high-income groups (Wall & Large, 2010), it is feasible that experiencing a greater dollar loss may also prompt reporting. Those who lose large amounts of money might be more inclined to report believing that law enforcement agencies would take their complaint more seriously and that the effort of reporting would be worthwhile.

The type of outlet where the counterfeit product was purchased might also influence victim reporting. The presence of reporting mechanisms at eBay and other online auction sites may increase victim awareness of counterfeit products and subsequent reporting (Chua et al., 2007; Gregg & Scott, 2006). In addition, victims may want to report seller misconduct at auction sites as a security measure for other buyers. Purchases made through other online retailers may be perceived as more risky (Forsythe, Liu, Shannon, & Gardner, 2006; Zhou, Dai, & Zhang, 2007). Consequently, consumers at these sites may be less inclined to make a complaint.

Having information about offenders might prompt victims to report. For example, individuals who check the reputation and sales history of online vendors may want to report their experience to ensure that they are caught or sanctioned in some way. Making multiple contacts with an offender may also prompt reporting. Victims who communicate with sellers through several channels, like e-mail, instant messaging, and cell phone calls, may have more details and better documentation of the fraud making it easier to support a complaint when filed, leading them to believe that reporting would be worthwhile.

It is unclear how the demographic characteristics of both victims and offenders may affect the overall likelihood of reporting. In fact, there is mixed evidence as to the influence of age, sex, marital status, and income on the risk of fraud victimization and reporting (Anderson, 2004; Baker & Faulkner, 2003; Blum, 1972; Holtfreter, Reisig, & Pratt, 2008; Schoepfer & Piquero, 2009; Titus, Heinzelman, & Boyle, 1995; Trahan, Marquart, & Mullings, 2005). This may be a consequence of the nature of fraud, as offenders generally attempt to victimize as many individuals as is possible in the course of a scheme (Titus et al., 1995). As a result, it is unclear how victim demographics affect reporting of product counterfeiting incidents. Similarly, it is unclear how suspect information influences reporting. The international scope of e-commerce allows offenders to target individuals anywhere around the world. As a result, victims may be more likely to report an incident involving a suspect within the United States believing that the investigation would be more promising than had it involved an international suspect.

As explained above, a primary purpose of our study is to examine the empirical relationship between these variables and various forms of victimization reporting. We aim to show how they affect, statistically, the odds of reporting when other factors are controlled.

Method

Data

In order to assess the likelihood of victim reporting, we obtained data from the IC3 for 2009 and 2010.² The IC3 is managed in part by the NW3C and the Federal Bureau of Investigation and serves as a primary reporting resource for victims of online crimes. Though the IC3 receives complaints on offenses ranging from child pornography to advance fee fraud victimization (Internet Crime Complaint Center, 2010), the most common reports involve consumer victimization via auction sites and other online retailers. We examined only reports of auction and non-auction fraud (defined below) relating to the delivery of fake products reported by U.S. victims. These types of victimization represent product counterfeiting or the intellectual property rights infringement of “material goods” or fake goods (International Organization for Standardization, 2010).

Citizens can file complaints directly with IC3 at their website (http://www.ic3.gov). According to NW3C staff (A. Wall-Parker and J. Lybarger, personal communication, August 20, 2012), complainants are directed to this website at the NW3C and FBI web pages, but they may also be referred to it by other entities, such as state and local law enforcement, eBay, PayPal, and the Federal Trade Commission. IC3 publicizes its services through several modalities, including maintaining a presence on the NW3C and FBI web pages, publishing annual reports, holding workshops for state and local law enforcement and victim advocacy groups, publicizing through NW3C’s various trainings for state and local law enforcement, presenting at various criminal justice and academic conferences, and fostering community awareness through presentations at venues such as schools and senior citizen centers. The IC3 website is continually updated to reflect current Internet crime trends and threats, and new products are frequently added to the website to further the public’s access to knowledge concerning online crime. IC3 complaint data can be viewed in “real time” and are regularly examined by in-house analysts who have the ability to assess complaints, analyze activity, and produce products that alert law enforcement, private sector, and the public to emerging Internet crimes and threats.

Despite the visibility of the IC3, this study focuses on U.S. victims to limit the potential selection bias and unreliability that might occur as a result of differences between U.S. and non-U.S. victims in their knowledge about the ability to report victimizations to IC3 or the willingness to do so (additionally, as discussed in the following section, the dependent variable measures whether the victim reported the incident to the Better Business Bureau [BBB], which is yet another U.S.-based institution with which U.S. victims would be more familiar). We also eliminated four complaints that had outliers for the variable dollar amount lost (i.e., $1,815,637; $300,000; $145,885; $130,000) as well as 30 complaints by victims whose nationality was unknown. Altogether, this reduced the final sample to 2,678 complaints.³

A notable limitation of these data is that they are entirely self-reported; as mentioned, victims may be directed to the IC3 by local law enforcement or online auction sites, though it is not otherwise advertised in a broad fashion online. As a consequence, these reports do not include all online consumer victimizations and present a substantive selection bias (these issues are discussed in greater detail later). It is unknown how well these reports represent the larger population of victims, thus our findings cannot be generalized beyond those who contact the IC3. Despite this limitation, the data permit the development of empirical and foundational lessons in an area where little empirical analysis exists.

Variables

Descriptive statistics for auction and non-auction complaints together are displayed in Table 1 (categorical variables) and Table 2 (count variables), and discussed in detail in the Results section. These tables include some variables not mentioned above but each is explained below in the context of the regression analyses. Also, because data were missing or unknown for several variables results are not always based on the full sample. The sample size for each statistic is noted in the tables.⁴

Table 1.

Descriptive Statistics of All Victims (Categorical Variables)

Variable	N	Frequency	%
U.S. suspect	1,797	1,305	72.6
Male victim	2,678	1,490	55.6
Victim under 30	2,678	823	30.7
Victim age	2,678
Under 20		113	4.2
20 to 29		710	26.5
30 to 39		657	24.5
40 to 49		599	22.4
50 to 59		419	15.6
Over 60		180	6.7
Auction victim	2,678	737	27.5
Third-party payment service	2,286	1,111	48.6
Method of payment	2,678
Credit card		610	22.8
Check/debit card		482	18.0
Cash		336	12.5
Wire transfer		295	11.0
Money order		126	4.7
Cashier’s check		59	2.2
Other		18	0.7
Mediums used by suspect	2,678
E-mail		960	35.8
Website		502	18.7
Telephone		445	16.6
Internet messaging		318	11.9
Mail		302	11.3
In-person contact		203	7.6
Other		180	6.7
Bulletin board		114	4.3
Wire		95	3.5
Fax		63	2.4
Chat room		48	1.8
Newsgroups		9	0.3
Initial contact unsolicited	2,529	648	25.6
Prior relationship with suspect	2,612	158	6.0
Researched suspect	2,572	826	32.1
Report to any entity	2,678	1,084	40.5
Reporting by entity	1,345
Individual/business		651	48.4
Law enforcement		409	30.4
Consumer protection agency		133	9.9
Better Business Bureau		113	8.4
Private attorney		39	2.9

Table 2.

Descriptive Statistics of All Victims (Count Variables)

	N	Minimum	Maximum	M	Median	SD
Mediums used by suspect	1,866	1	8	1.73	1.00	0.95
Dollar amount lost	2,678	0	100,000	1,328.66	164.00	5,933.52
Reporting depth	1,084	1	5	1.24	1.00	0.58

Dependent variables

In order to assess the likelihood or reporting a victimization experience, we recoded the dependent variable, report to any entity (1 = yes and 0 = no), to measure whether or not the victim reported the incident to any of the following entities identified in the IC3 complaint form: the individual/business that victimized him or her (i.e., the suspect), the BBB, a consumer protection agency, a private attorney, or the police/other law enforcement. Nearly 60% of complainants did not report their victimization to any of these entities. On average, victims who did report contacted just one of these entities. Of the total reports made (N = 1,345), almost half were against the individual or business who sent the fraudulent product (often treated as the suspect by IC3), which was the most among all the entities. Less than a third of reports were made to law enforcement. Finally, reports to a consumer protection agency, the BBB, and a private attorney were much less common, each comprising less than 10% of all reports, respectively.

Independent variables

Currently, there is little theory to guide the analysis of product counterfeiting, which makes it difficult to state and test formal hypotheses. Therefore, in this article, we take an exploratory approach and describe several variables available in the IC3 data and examine their influence on victim reporting. Below, we discuss how each variable is measured and the expected relationships based on the literature discussed earlier.

From the original data, we created four binary predictors. First, U.S. suspect (1 = yes and 0 = no) broadly measures the location of the suspect. In general, we expect that being victimized by U.S. suspects would prompt reporting because victims may think they are more likely to be pursued by authorities compared to international suspects. Law enforcement agencies may also be more willing to investigate these offences due to the likelihood of clearance by arrest if both the victim and the offender are from the United States. Second, based on Dolan (2004), we included a measure of victim age—victim under 30 years of age (1 = yes and 0 = no). He found that victims reporting auction fraud victimization to the NW3C were typically over 30 years old, so we expect that being younger is associated with less reporting in general. Third, prior relationship with suspect (1 = yes and 0 = no) measures whether the victim knew the suspect in some capacity prior to the incident (in the IC3 complaint form, known suspects include someone at work or school, a business acquaintance, a relative, a friend or neighbor, an online acquaintance, or someone else met or known). We expect this variable to be positively related to reporting because being able to specifically identify a suspect helps substantiate a complaint. Fourth, auction victim (1 = yes and 0 = no, non-auction victim) accounts for the type of website from which victims purchased counterfeit products. For auction victims, this includes Internet auction houses like eBay and Half.com where goods are purchased through a bid-style format. Non-auction victims were from any other online consumer goods outlet where products are purchased at a fixed price such as Amazon or Craigslist. Refining our analyses by victimization type is consistent with the crime prevention and victimology literatures which emphasize the need to understand specific types of victims (see Clarke & Eck, 2005). This is why auction and non-auction victims are analyzed separately in the descriptive analysis. For reasons mentioned in the literature review, we expect (in the regression analyses) that auction victims are more likely to report their victimization.

As seen in the descriptive analyses, we examined all categories of payment methods and communication mediums from the IC3 form. However, we also transformed the categorical communication mediums variable into a count variables for the regression analyses: total number of mediums used by suspect to communicate with the victim in the course of the incident (e.g., bulletin board, chat room, e-mail, fax, in-person contact, Internet messaging, mail, newsgroups, telephone, website, wire, and other medium).⁵ We expect this variable to be positively associated with reporting—having more evidence of a transaction from records of communication would provide greater support for a complaint and might encourage reporting.

Most other predictors did not require transformation. We operationalized victim sex as male victim (1 = yes and 0 = no). Again, we include this variable based on Dolan (2004) and expect that males are more likely to report in general. We included a measure for initial contact unsolicited (1 = yes and 0 = no) to assess the effect of having an unsolicited or uninvited initial contact with the suspect on reporting. Certain forms of fraud, such as advance fee fraud schemes are generated through unsolicited initial contact (Holt & Graves, 2007; Wall, 2004), though there is little quantitative research related to this variable. We therefore do not hypothesize the direction of the relationship but rather wish to test whether a relationship exists. We included the variable researched suspect (1 = yes and 0 = no), which measures whether or not the victim conducted any research on the individual or business prior to the incident. We expect this to be positively related to reporting because having greater knowledge of the suspect may prompt reporting. Finally, to determine how the economic impact of the incident affects reporting, we included dollar amount lost (a continuous measure of the victim’s reported monetary loss). Again, we expect a positive relationship with reporting. Just over 14% of victims reported a zero dollar loss for this variable, while almost 25% of victims reported losing less than 50 dollars. Additionally, 26 victims reported a loss of less than 10 dollars. We used the natural log of the dollar lost variable in the logistic regressions to reduce its skewness and scale. Because the natural log of zero values in count data cannot be taken, we used the conventional procedure of adding the constant of 0.5 to each value to make them all nonzero (see McDonald, 2009).

Analysis

We conducted three analyses to assess the factors related to counterfeit product victimization. First, we describe the characteristics and reporting of all incidents together. Second, for reasons described above, we examine these same factors but broken down by auction and non-auction victimization. Finally, we conduct separate binary logistic regressions to predict victim reporting in general and also reporting to either law enforcement or the suspect, specifically (auction and non-auction victims are combined in each model). We evaluated each model for its overall ability to statistically predict reporting, the substantive and statistical significance of predictor coefficients, and the ability of the model to correctly predict observations.⁶

Results

Based on complaints made to the IC3, it was evident that most incidents involve non-auction victimizations. Nearly three fourths of known suspects are from the United States. Most victims are male and about 70% were aged 30 or older. Of specific age groups, however, most victims are between 20 and 29 years old. Just over half of victims did not use a third-party online payment service (e.g., PayPal, BidPay, or Escrow). Credit cards and check/debit cards are the most commonly used method of payment. Suspects contacted victims in a variety of ways but most often through e-mail and websites. In addition, suspects and victims communicated through an average of less than two mediums over the course of the incident. Only a quarter of incidents involved initial contact by suspects that was unsolicited or uninvited. Given that incidents are online transactions, it is not surprising that very few victims had a relationship with the suspect prior to the incident.

Descriptive statistics by complaint type are shown in Table 3 (categorical variables) and Table 4 (count variables). Both groups are usually male and victimized by U.S. suspects. Non-auction victims were slightly younger and most were between 20 and 29 years old, while auction victims were typically between 40 and 49 years old. Using third-party payment services was more common among auction victims. This is expected, given that this type of payment is available and often preferred at popular online auction websites like eBay. However, both groups most often use credit and check/debit cards to pay for purchases. Suspects prefer e-mail to contact both groups but making unsolicited initial contact is more common in non-auction incidents. More auction victims conduct research on suspects prior to the incident. This is also expected because buyers at auction sites can easily review seller reviews and feedback before making a purchase.

Table 3.

Descriptive Statistics by Type of Victimization (Categorical Variables)

	Auction			Nonauction
Variable	N	Frequency	%	N	Frequency	%
U.S. suspect	607	477	78.6	1,190	828	69.6
Male victim	737	435	59.0	1,941	1,055	54.4
Victim under 30	737	189	25.6	1,941	634	32.7
Victim age	737			1,941
Under 20		28	3.8		85	4.4
20 to 29		161	21.8		549	28.3
30 to 39		171	23.2		486	25.0
40 to 49		173	23.5		426	21.9
50 to 59		142	19.3		277	14.3
Over 60		62	8.4		118	6.1
Third-party payment service	687	566	82.3	1,599	545	34.1
Method of payment	737			1,941
Credit card		192	26.1		418	21.5
Check/debit card		130	17.6		352	18.1
Cash		31	4.2		305	15.7
Wire transfer		78	10.6		217	11.2
Money order		29	3.9		97	5.0
Cashier’s check		10	1.4		49	2.5
Other		6	0.8		12	0.6
Mediums used by suspect	737			1,941
E-mail		309	41.9		651	33.5
Website		109	14.8		393	20.2
Telephone		95	12.9		350	18.0
Internet messaging		98	13.3		220	11.3
Mail		119	16.1		183	9.4
In-person contact		13	1.8		190	9.8
Other		69	9.4		111	5.7
Bulletin board		10	1.4		104	5.4
Wire		24	3.3		71	3.7
Fax		17	2.3		46	2.4
Chat room		5	0.7		43	2.2
Newsgroups		1	0.1		8	0.4
Initial contact unsolicited	688	133	19.3	1,841	515	28.0
Prior relationship with suspect	720	35	4.9	1,892	123	6.5
Researched suspect	696	315	45.3	1,876	511	27.2
Report to any entity	737	328	44.5	1,941	756	38.9
Reporting by entity	407			938
Individual/business		254	62.4		397	42.3
Law enforcement		86	21.1		323	34.4
Consumer protection agency		33	8.1		100	10.7
Better Business Bureau		19	4.7		94	10.0
Private attorney		15	3.7		24	2.6

Table 4.

Descriptive Statistics by Type of Victimization (Count Variables)

	Auction (Nonauction)
	N	Minimum	Maximum	M	Median	SD
Mediums used by suspect	511	1	7	1.69	1.00	0.98
	(1,355)	(1)	(8)	(1.75)	(1.00)	(0.93)
Dollar amount lost	737	0	60,000	1,121.36	223.00	3648.60
	(1,941)	(0)	(100,000)	(1,407.37)	(140.00)	(6,596.10)
Reporting depth	328	1	4	1.24	1.00	0.58
	(756)	(1)	(5)	(1.24)	(1.00)	(0.58)

The majority of complainants in both groups did not report the incident to any entity, but auction victims reported slightly more often. For both groups, reports against the suspect are most common, followed by law enforcement, consumer protection agencies, the BBB, and private attorneys. However, these proportions differ considerably between auction and non-auction victims. Specifically, nearly two thirds of reports by auction victims are made against the suspect compared to just 42% of reports by non-auction victims. Furthermore, a much greater percentage of reports are made to law enforcement by nonauction than auction victims (34% vs. 21%). Both groups are similar on count variables (see Table 4), but auction victims lose slightly more money per transaction with a median of $233 versus $140.

Assessing the Factors Related to Counterfeit Victimization Reporting

Given the relationships noted between reporting behavior for auction and non-auction victims, we estimated a series of binary logistic regression models to identify factors associated with the odds of reporting. In order to conduct these analyses, we performed multiple imputation procedures using the standard of five imputations and interpreted the pooled logistic regression estimates. This procedure is superior to simply deleting cases with missing variables because it does not reduce the sample size, takes advantage of information that is provided, and reduces selection bias that can occur by only retaining only those cases for which full information is available (those cases could be unrepresentative of the entire sample; Schafer & Olsen, 1998). Additionally, for each of the reporting options (any entity, law enforcement, and suspect), we present the results for two separate regression models. The “full” models include nine predictors. We removed two variables in the reduced models: U.S. suspect and mediums used by suspect.

These variables are each missing roughly one third of their values, requiring a large proportion of data to be imputed. It is not surprising that the location of suspects is often unknown, given that transactions occurred online. It is less clear why missing data were prevalent for the other variables. The remaining variables are more stable. The variable with the next highest missing values (initial contact unsolicited) requires only about 5% of its values imputed. As a sensitivity analysis, we use the reduced models to examine whether the effects of other variables remain stable when those with high missing values are excluded from the model.

Table 5 presents the results of the models predicting victim reporting to any entity. Based on the full model, being male and less than 30 years old coincides with not reporting. The odds of reporting to any entity increase with the dollar amount lost, as expected. Also consistent with expectations, there is a positive association between reporting and the victim having a relationship with the suspect individual or business before victimization, researching them, and having contact with them via more mediums. This suggests that having a greater knowledge of or familiarity with the suspect may prompt reporting. In fact, the odds ratio of prior relationship is notable as it indicates the odds of reporting to any entity essentially double when the victim has a prior relationship with the suspect individuals or business. Finally, being an auction (as opposed to nonauction) victim, the nature of initial contact between victim and suspect nationality are not associated with the odds of reporting to any entity.

Table 5.

Binary Logistic Models Predicting Any Victim Reporting

Predictor	Odds ratios
Predictor	Full model	Reduced model
Male victim	.839*	.853
Victim under 30	.757**	.755**
Auction victim	1.064	1.046
Dollar amount lost (ln)	1.182**	1.191**
Initial contact unsolicited	1.205	1.163
Prior relationship with suspect	1.990**	1.872**
Researched suspect	1.261*	1.286**
U.S. suspect	1.070
Mediums used by suspect	1.170**
N	2,678	2,678

Note. *p < .05. **p < .01.

For the variables included in the reduced model, all of the findings hold except for being a male victim, which is not statistically related to reporting in this model. This suggests that the role of sex in reporting is somehow influenced by the variables containing larger amounts of missing data (for U.S. suspect and mediums used by suspect).

We created additional logistic regression models to identify the correlates of reporting specifically to law enforcement organizations (see Table 6) and to suspect individuals or businesses (see Table 7). As seen in Table 6, auction victims (as opposed to nonauction) are less likely to report their victimization to law enforcement, but the odds of this kind of reporting increase with the value lost and when the initial contact by suspect was unsolicited. In addition, suspect nationality has a substantive relationship with reporting to law enforcement. According to the model, as expected victims are about 2½ times more likely to report their victimization to law enforcement if the suspect has U.S. nationality.

Table 6.

Binary Logistic Models Predicting Victim Reporting to Law Enforcement

Predictor	Odds ratios
Predictor	Full model	Reduced model
Male victim	1.117	1.260
Victim under 30	0.944	0.937
Auction victim	0.554**	0.591**
Dollar amount lost (ln)	1.367**	1.392**
Initial contact unsolicited	1.416*	1.268
Prior relationship with suspect	3.457**	2.907**
Researched suspect	0.694**	0.691**
U.S. suspect	2.603**
Mediums used by suspect	1.140**
N	2,678	2,678

Note. *p < .05. **p < .01.

Table 7.

Binary Logistic Models Predicting Victim Reporting to Suspect

Predictor	Odds ratios
Predictor	Full model	Reduced model
Male victim	.714**	.692**
Victim under 30	.760**	.761**
Auction victim	1.857**	1.731**
Dollar amount lost (ln)	1.074**	1.082**
Initial contact unsolicited	.947	.940
Prior relationship with suspect	.514**	.493**
Researched suspect	1.633**	1.692**
U.S. suspect	.711*
Mediums used by suspect	1.215**
N	2,678	2,678

Note. *p < .05. **p < .01.

Having a greater knowledge of, or familiarity with, a suspect influences law enforcement reporting somewhat differently than general reporting to any entity. While more mediums of communication between victim and suspect similarly increased the odds of both types of reporting, an existing relationship with the suspect more than triples the odds of reporting to law enforcement but having researched the suspect beforehand reduces the odds of such reporting. Unlike the general reporting model, victim sex and age are unrelated to the odds of reporting to law enforcement. The reduced model supports the findings of the full model with the exception of having unsolicited initial contact. For reporting to law enforcement, this factor appears to be associated with the variables containing larger amounts of missing data.

Table 7 provides the results of the model predicting reporting to the suspect individual or business. Being male and less than 30 years of age is associated with reduced odds of this type of reporting. The amount lost is positively related to the odds of reporting and being an auction victim almost doubles the odds of reporting to the suspect. Interestingly, being victimized by a U.S. suspect increased the odds of reporting to law enforcement but reduced the odds of reporting to the actual suspect. It is also notable how familiarity and interactions with the suspect influence types of reporting differently. Although the number communication mediums similarly increased the odds of reporting to both law enforcement and the suspect, having a prior relationship with the suspect more than triples the odds of reporting to law enforcement but essentially cuts in half the odds of reporting to the suspect. Conversely, having researched the suspect before the transaction reduces the odds of reporting to law enforcement, yet substantively increases the odds of reporting to the suspect. The form of initial contact is not associated with reporting victimizations to suspects, unlike the law enforcement model where the relationship was positive. Finally, for suspect reporting, the reduced form model supported all of the findings in the full model.

Discussion and Conclusion

The sale and distribution of counterfeit products through online retailers has increased substantially over the last decade (Wall & Large, 2010). Few criminologists, however, have considered the basic characteristics of this type of victimization let alone the factors that influence reporting such incidents. To address this gap, we examined complainant data from the 2009 and 2010 IC3 reports to describe these incidents and identify the demographic and situational correlates of reporting. Collectively, the logistic regression models revealed several important factors associated with victim reporting, both at odds and consistent with expectations.

Being male and less than 30 years old reduces the odds of any reporting and reporting to the suspect specifically, but these characteristics are unrelated to law enforcement reporting. Online auction victims are also less likely to report their victimizations to law enforcement but are more likely to report victimization to the suspects themselves. As expected, losing more money is consistently and positively related with all forms of reporting. The form of initial contact only affects reporting to law enforcement such that when it is unsolicited the odds of reporting increase.

The role of victim–suspect familiarity also influences reporting behavior. An existing prior relationship with the suspect increases the odds of reporting generally and to law enforcement specifically but reduces the odds of reporting to the actual suspect individual or business. By contrast, a victim who researched the suspect prior to the purchase is more likely to report the victimization to any entity generally and the suspect specifically but is less likely to report the crime to law enforcement. The number of communication mediums used is consistently associated with increased odds of all types of reporting. Finally, consumers victimized by U.S. nationals are more likely to complain to law enforcement but less likely to complain to the suspects themselves. This divergent finding may explain why nationality of the suspect is statistically unrelated to reporting to any entity.

The descriptive results suggest that auction and non-auction product counterfeiting victims are minimally different. The most substantive difference is that non-auction victims use less secure forms of payment. The lack of substantive differences between groups supports the notion that fraudsters attempt to target as many groups as possible in the course of their scams (Holtfreter et al., 2008; Titus et al., 1995). In addition, this study validates the mixed findings on the demographic composition of fraud victims (see Holtfreter et al., 2008 for discussion).

These findings, however, must be considered in light of the limitations of our data. This major limitation stems largely from our focus on an emerging crime where little data and research currently exist. A major hurdle in studying product counterfeiting is the fact that reliable data are either nonexistent or unavailable to researchers. Therefore, it is necessary to make use of such data when they become available, even when they may not be ideal. This is the situation with this article and the IC3 complainant data.

We recognize that our sample likely represents only a small proportion of all online product counterfeiting victimizations and is limited by selection issues, which raise questions about the generalizability of study findings. Many victims simply may not know the IC3 exists, and if they do, they may choose not to file a complaint or instead report only to another entity, like the police (so the incident is captured elsewhere but not by the IC3). Reporting to the IC3 may also be influenced by Internet accessibility, although if the victim made the initial purchase online this effect may be minimal. Of course, individuals would not report a victimization if they did not realize they received a counterfeit item.

Another limitation of the data is that we cannot detect the presence of repeat victims. However, research suggests that repeat victims are unlikely to report their victimization to the police (see van Dijk, 2001), so there is reason to suspect that victims are also less likely to file multiple complaints with the IC3. Furthermore, individuals may alter their behavior after an initial victimization (e.g., they no longer shop online or buy from different merchants), which could reduce the occurrence of future victimizations and the need to report them. Together, these factors suggest that repeat victims could comprise a relatively small proportion of the data. Either way, our work should simply be interpreted as instances of victimization and not individual victims.

In sum, the lack of knowledge and research on product counterfeiting reflects the absence of valid and reliable data on this crime. Accordingly, at this time it is not reasonable to expect that attempts to empirically examine product counterfeiting, such as this study, will be based on a data set and sample that we would generally interpret as ideal. However, we acknowledge the issues inherent in the IC3 complainant data and therefore again urge readers to consider the findings of this study cautiously. Furthermore, in the discussion below, we also avoid making policy recommendations and instead focus on the potential contribution of our findings for advancing theory and directing future research.

Currently, criminology does not provide a strong theoretical framework for understanding the nature of product counterfeiting and counterfeiters or the risk factors associated with victimization. Instead, more recent theoretical developments in this area have occurred in fields like marketing and economics (for an example of the former see Wilcox, Hyeong, & Sankar, 2009, and for the latter see Naylor, 2003). Consequently, we lack a clear understanding of product counterfeiting from a criminological perspective. This may stem from the lack of reliable data on this form of victimization from official and self-report sources (see National White Collar Crime Center, 2004; Olsen, 2005; Piquero, 2005; USGAO, 2010a). Piquero (2005, p. 57), for instance, suggests that little is known about the “causes and correlates of intellectual property theft” and “what is known seems to raise more questions than answers.”

To improve our knowledge, criminology may benefit from substantive exploratory studies that present basic characteristics and correlates of online product counterfeiting victimization and reporting. Our findings provide early insight into several important aspects of this crime, including target selection, the offender–victim dyad, and online crime “places” (see Eck, 2002 and Sherman, Gartin, & Buerger, 1989 for a discussion of the importance of place and crime and Newman & Clarke, 2003 for virtual places, specifically).

This study also provides guidance for future research on product counterfeiting. First, in just a few years of data, there were thousands of individuals who considered their victimization serious enough to file a complaint with the IC3 (at the very least), with some also reporting to law enforcement and other authorities. Although there are questions about the generalizability of complainants, this suggests that online consumers are concerned about product counterfeiting and seek resolution after victimization, so steps could be taken to encourage reporting incidents to authorities who are in a position to help, regardless of their familiarity with the suspect and the cost of the item.

Future research is needed with quantitative and qualitative data sources to fully explore how seriously online consumers perceive the threat of product counterfeiting and what factors increase or decrease the likelihood of reporting victimization. At the same time, research should assess the extent to which law enforcement, especially local police agencies, prioritizes product counterfeiting and what, if anything, they need to bolster their response when it is reported. Recent evidence suggests that state and local police actively combat product counterfeiting (Heinonen & Wilson, 2012), but this finding is limited to Michigan-related incidents and does not fully clarify law enforcement’s perception of this crime or their capacity to address it. It is also necessary to consider how different agencies respond to complaints about counterfeiting.

Second, given that some complainants reported their victimization to multiple entities, there are opportunities for different authorities to collaborate in response to victim complaints. This is likely challenging, but required due to the nature of counterfeiting affecting police, corporations, and other agencies depending on the type of product counterfeited. In particular, these strategies may be useful for incidents involving international suspects, which was the case in nearly 30% of the incidents examined here. Future research is needed to identify ways to best create and promote such collaborative efforts, whether through resources, training, or the distribution of promising practices or some other type of information-sharing mechanism.

Finally, given that almost 30% of incidents occurred at auction websites, future research should examine how these websites respond to reported incidents of product counterfeiting. Unlike non-auction purchases that may occur at temporary, rogue websites operated by criminals, auction purchases likely occur at stable, reputable websites that monitor buyers, sellers, and transactions. Therefore, researchers could solicit auction websites for data related to, among other things, the types of counterfeit products sold, methods of payment, seller registrations and profile changes, seller locations, and descriptions of counterfeit products auction listings. At minimum, researchers could examine how major online outlets and retailers, like eBay, PayPal, and Amazon, communicate the threat of product counterfeiting to consumers. Such empirical research could expand our understanding of the nature of product counterfeiting on and off-line, as well as identify the factors that influence reporting and the role of criminology in better understanding this crime.

Footnotes

Acknowledgment

We would like to thank NW3C for providing the data for this study.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Albanese

J. S.

(2011). Transnational crime and the 21st century: Criminal enterprises, corruption, and opportunity. New York, NY: Oxford University Press.

Anderson

(2004). Consumer fraud in the United States: An FTC survey. Washington, DC: Federal Trade Commission. Retrieved from www.ftc.gov/reports/consumerfraud/040805confraudrpt.pdf

Baker

W. E.

Faulkner

R. R.

(2003). Diffusion of fraud: Intermediate economic crime and investor dynamics. Criminology, 41, 1173–1206.

Balsmeier

Bergiel

B. J.

Viosca

R. C.

Jr (2004). Internet fraud: A global perspective. Journal of E-Business, 4, 1–12.

Bass

Hoeffler

(1992). Telephone-based fraud: A survey of the American public. New York, NY: Louis Harris and Associates.

Bate

(2008). Making a killing: The deadly implications of the counterfeit drug trade. Washington, DC: American Enterprise Institute Press.

Baum

(2006). First estimates from the National Crime Victimization Survey: Identity theft, 2004. Washington, DC: U.S. Department of Justice, Office of Justice Statistics. Retrieved from http://www.ojp.usdoj.gov/bjs/pub/pdf/it04.pdf

Blum

R. H.

(1972). Deceivers and deceived: Observations on confidence men and their victims, informants and their quarry, political and industrial spies and ordinary citizens. Springfield, IL: Charles C. Thomas.

Chow

D. C.

(2003). Organized crime, local protectionism, and the trade in counterfeit goods in China. China Economic Review, 14, 473−484.

10.

Chua

C. E. H.

Wareham

Robey

(2007). The role of online trading communities in managing Internet auction fraud. MIS Quarterly, 31, 750–781.

11.

Clarke

R.V.

Eck

J. E.

(2005). Crime analysis for problem solvers in 60 small steps. Washington, DC: US Department of Justice and the Center for Problem Oriented Policing.

12.

Dolan

K. M.

(2004). Internet auction fraud: The silent victims. Journal of Economic Crime Management, 2, 1–22.

13.

Eck

J. E.

(2002). Preventing crime at places. In Sherman

L. W.

Farrington

Welsh

MacKenzie

D. L.

(Eds.), Evidence-based crime prevention (pp. 241–294). New York, NY: Routledge.

14.

Federal Trade Commission. (2011). Consumer sentinel network data book for January–December 2010. Retrieved from http://ftc.gov/sentinel/reports/sentinel-annual-reports/sentinel-cy2010.pdf

15.

Forsythe

Liu

Shannon

Gardner

L. C.

(2006). Development of a scale to measure the perceived benefits and risks of on-line shopping. Journal of Interactive Marketing, 20, 55–75.

16.

Freeman

Sidhu

N. D.

Montoya

(2007, February). A false bargain: The Los Angeles County economic consequences of counterfeit products. Report. Los Angeles County Economic Development Corporation.

17.

Gartner, Inc. (2003). Gartner says marketers must differentiate e-mail marketing from spam: Spam will account for 60 percent of e-mail by end of 2004. Retrieved from http://www.gartner.com/press_releases/pr29sept2003a.html

18.

Gregg

D. G.

Scott

J. E.

(2006). The role of reputation systems in reducing on-line auction fraud. International Journal of Electronic Commerce, 10, 95–120.

19.

Heinonen

J. A.

Wilson

J. M.

(2012). Product counterfeiting: Evidence-based lessons for the State of Michigan. East Lansing, MI: A-CAPPP Paper Series.

20.

Holt

T. J.

(2003). Examining a transnational problem: An analysis of computer crime victimization in eight countries from 1999 to 2001. International Journal of Comparative and Applied Criminal Justice, 27, 199–220.

21.

Holt

T. J.

Graves

D. C.

(2007). A Qualitative analysis of advanced fee fraud schemes. The International Journal of Cyber-Criminology, 1, 137–154.

22.

Holtfreter

Reisig

M. D.

Blomberg

T. G.

(2006). Consumer fraud victimization in Florida: An empirical study. Saint Thomas Law Review, 18, 761–789.

23.

Holtfreter

Reisig

M. D.

Pratt

T. C.

(2008). Low self-control, routine activities, and fraud victimization. Criminology, 46, 189–220.

24.

International Anti-Counterfeiting Coalition. (2005, January). The negative consequences of international intellectual property theft: Economic harm, threats to the public health and safety, and links to organized crime and terrorist organizations. Report. Washington, DC: International Anti-Counterfeiting Coalition.

25.

ICC Counterfeiting Intelligence Bureau. (1997). Countering counterfeiting: A guide to protecting and enforcing intellectual property rights. Paris, France: International Chamber of Commerce Publishing.

26.

International Organization for Standardization. (2010). Fraud countermeasures and controls. Retrieved from http://www.iso.org/iso/standards_development/technical_committees/other_bodies/iso_technical_committee.htm?commid=580925

27.

Internet Crime Complaint Center. (2010). IC3 2010 Internet crime report. Retrieved March 24, 2010, from http://www.ic3.gov/media/annualreport/2010_IC3Report.pdf

28.

Internet Crime Complaint Center. (2011). IC3 2011 Internet crime report. Retrieved June 15, 2012, from http://www.ic3.gov/media/annualreport/2011_IC3Report.pdf

29.

James

(2005). Phishing exposed. Rockland, MA: Syngress.

30.

Kelly

(2005). Illicit trafficking: A reference handbook. Santa Barbara, CA: ABC-CLIO.

31.

Lai

K. K.

Zaichkowsky

J. L.

(1999). Brand imitation: Do the Chinese have different views? AsiaPacific Journal of Management, 16, 179–192.

32.

Liang

B. A.

(2006). Fade to black: Importation and counterfeit drugs. American Journal of Law & Medicine, 32, 279−323.

33.

Mason

Benson

(1996). The effect of social support on fraud victim’s reporting behavior: A research note. Justice Quarterly, 13, 511−523.

34.

McDonald

J. H.

(2009). Handbook of biological statistics. Baltimore, MD: Sparky House Publishing.

35.

Motor & Equipment Manufacturers Association (MEMA). (2009, January). Understanding the flow of counterfeit and gray market goods through the U.S. automotive and commercial vehicle parts marketplace. Report. Washington, DC: Motor and Equipment Manufacturers Association.

36.

Nasheri

(2005). Addressing the global scope of intellectual property crimes and policy initiatives. Trends in Organized Crime, 8, 79−108.

37.

National White Collar Crime Center. (2004). Intellectual property and white-collar crime: Report of issues, trends, and problems of future research. Washington, DC: US Department of Justice.

38.

Naylor

R. T.

(2003). Towards a general theory of profit-driven crimes. British Journal of Criminology, 43, 81–101.

39.

Nerenberg

(2000). Forgotten victims of financial crime and abuse: Facing the challenge. Journal of Elder Abuse and Neglect, 12, 49–73.

40.

Newman

Clarke

(2003). Superhighway robbery: Preventing e-commerce crime. Cullompton, England: Willan Press.

41.

Organization for Economic Cooperation and Development. (2007). The economic impact of counterfeiting and piracy. Report. Author. Retrieved from http://www.oecd.org/dataoecd/13/12/38707619.pdf

42.

Olsen

(2005). Counterfeiting and piracy: Measurement issues. Retrieved from http://www.oecd.org/dataoecd/42/44/35651123.pdf

43.

Phillips

(2005). Knockoff: The deadly trade in counterfeit goods: The true story of the world’s fastest growing crime wave. London, England: Kogan Page.

44.

Piquero

N. L.

(2005). Causes and prevention of intellectual property crime. Trends in Organized Crime, 8, 40–61.

45.

Raustiala

Sprigman

(2005). The knockoff ecomony: How imitation sparks innovation. Oxford, England: Oxford University Press.

46.

Rebovich

D. J.

Layne

(2000). National public survey on White collar crime. Morgantown, WV: National White Collar Crime Center.

47.

Schafer

J. L

Olsen

M. K.

(1998). Multiple imputation for multivariate missing-data problems: A data analyst’s perspective, Multivariate Behavioral Research, 33, 545–571.

48.

Schoepfer

Piquero

N. L.

(2009). Studying the correlates of fraud victimization and reporting. Journal of Criminal Justice, 37, 209–215.

49.

Sherman

Gartin

Buerger

(1989). Hot spots of predatory crime: Routine activities and the criminology of place. Criminology, 27, 27–55.

50.

Shichor

Sechrest

Doocy

(2000). Victims of investment fraud. In Pontell

Shichor

(Eds.), Contemporary issues in crime and criminal justice: Essays in honor of Gilbert Geis (pp. 87−96). Upper Saddle River, NJ: Prentice Hall.

51.

Thompson

J. W. C.

(2004). Bootleg billions: The impact of the counterfeit goods trade on New York City. Report. New York: City of New York Office of the Comptroller.

52.

Titus

R. M.

Heinzelman

Boyle

J. M.

(1995). Victimization of persons by fraud. Crime and Delinquency, 41, 54−72.

53.

Trahan

Marquart

J. W.

Mullings

(2005). Fraud and the American dream: Toward an understanding of fraud victimization. Deviant Behavior, 26, 601–620.

54.

United States General Accountability Office. (2010a, April). Intellectual property: Observations on efforts to quantify the economic effects of counterfeit and pirated goods (Report no. GAO-10-423). Washington, DC: Author.

55.

United States General Accountability Office. (2010b, October). Intellectual property: Agencies progress in implementing recent legislation, but enhancements could improve future plans (Report no. GAO-11-39). Washington, DC: United States General Accountability Office.

56.

van Dijk

J. J. M.

(2001). Attitudes of victims and repeat victims toward the police: Results of the International Crime Victims Survey. In Clarke

R. V.

Pease

(Eds.), Crime prevention studies (Vol. 12, pp. 27–52). Monsey, NY: Criminal Justice Press.

57.

Van Wyk

Mason

K. A.

(2001). Investigating vulnerability and reporting behavior for consumer fraud victimization: Opportunity as a social aspect of age. Journal of Contemporary Criminal Justice, 17, 328–345.

58.

Vaughan

Carlo

(1976). The appliance repairman: A study of victim-responsiveness and fraud. Journal of Research in Crime and Delinquency, 1975, 153–161.

59.

Wall

(2004). Digital realism and the governance of spam as cybercrime. European Journal on Criminal Policy and Research, 10, 309–335.

60.

Wall

D. S.

Large

(2010). Locating the public interest in policing counterfeit luxury fashion goods. British Journal of Criminology, 50, 1094–1116.

61.

Wilcox

Hyeong

M. K.

Sankar

(2009). Why do consumers buy counterfeit luxury brands? Journal of Marketing Research, 46, 247–259.

62.

Wood

P. A.

(2004). Spammer in the works: Everything you need to know about protecting yourself and your business from the rising tide of unsolicited “spam” e-mail. Retrieved from http://www.construct-it.org.uk/pages/sources/A%20spammer%20in%20the%20works.pdf

63.

Zhou

Dai

Zhang

(2007). Online shopping acceptance model: A critical survey of consumer factors in online shopping. Journal of Electronic Commerce Research, 8, 41–62.