Public Perceptions of White Collar Crime Culpability: Bribery, Perjury, and Fraud
PUBLIC PERCEPTIONS OF WHITE COLLAR CRIME CULPABILITY: BRIBERY, PERJURY, AND FRAUD
STUART P. GREEN 0 1
MATTHEW B. KUGLER 0 1
0 Copyright © 2012 by Stuart P. Green and Matthew B. Kugler. This article is also available at
1 Professor of Law and Justice Nathan L. Jacobs Scholar, Rutgers School of Law-Newark. For help in formulating our study instrument, we are grateful for comments from Sara Sun Beale, Pamela Bucy Pierson, Peter Henning, and Christine Parker. For their comments and questions, we also thank participants at the Duke Law and Contemporary Problems symposium conference, Adjudicating the Guilty Mind
We are accustomed to thinking of “crime” as involving the most blameworthy and antisocial sorts of conduct in which citizens can engage, conduct that is clearly and unambiguously more wrongful than conduct that is not criminal. But the reality is more complex, especially when we look at certain kinds of “white collar” behavior. One of us (Green) has previously undertaken an in-depth investigation of the underlying moral concepts that distinguish white collar crime from “merely aggressive behavior.”1 This work attempted to differentiate, for example, between criminal fraud and “sharp dealing,” insider trading and “savvy investing,” bribery and “horse trading,” tax evasion and “tax avoidance,” extortion and “hard bargaining,” witness tampering and “witness preparation,” and perjury and “wiliness on the witness stand.”2 Such analysis often depended on fairly fine-grained distinctions in moral reasoning. The problem is that the ability of criminal law to stigmatize, to achieve legitimacy, and to gain compliance ultimately depends on the extent to which it enjoys moral credibility and recognition in the broader lay community.3 1. See STUART P. GREEN, LYING, CHEATING, AND STEALING: A MORAL THEORY OF WHITE COLLAR CRIME (2006). We recognize that the concept of “white collar” crime has been highly contested in the academic literature, and that its use varies significantly even in common discourse. For present purposes, we adhere to the “moral” conception of white collar crime previously developed by Green, id. 9-29, according to which the term “white collar crime” refers to a loose collection of criminal offenses and related conduct distinguished by distinctive forms of harm and victimization, distinctive forms of wrongfulness, and a distinctive role for mens rea. 2. See id. 3. We have previously discussed this point in Stuart P. Green & Matthew B. Kugler, Community Perceptions of Theft Seriousness: A Challenge to Model Penal Code and English Theft Act Consolidation, 7 J. EMPIRICAL LEGAL STUDIES 511, 515-17 (2010). See also Stuart P. Green, Taking it to the Streets, 89 TEX. L. REV. SEE ALSO 61 (2011), http://www.texaslrev.com/seealso/ vol/89/responses/green.
If legally significant distinctions between fraud and non-fraud—or perjury and
non-perjury—can be discerned only through abstract philosophical reasoning, it
is reasonable to wonder whether the public will lend these distinctions the
moral weight required for the law to be effective and legitimate.
This paper seeks to determine the extent to which the lay public’s moral
intuitions parallel the law in distinguishing between white collar crime and
related non-crime by focusing on three domains of conduct: (1) bribery and
gratuities; (2) perjury and false statements; and (3) fraud. These types of
conduct are of practical significance and reflect the kind of moral ambiguity that
is characteristic of white collar crime. This paper examines each category to
determine where the lay public would draw the line between criminality and
non-criminality; and, where such conduct is regarded as criminal, how it would
be graded. The analysis aims to identify the extent to which public perceptions
are consistent or inconsistent with current law.
Our studies found that lay persons, in general, are comfortable making fairly
fine-grained distinctions regarding the law of white collar crime. In some cases,
the distinctions made by our respondents were consistent with current law; this
should lend weight to the view that the law in these areas draws distinctions in
the appropriate places. Participants in the fraud study, for example, were
comfortable distinguishing between misrepresentations that went to the heart of
the bargain and those that were extraneous. Elsewhere, however, we found
significant divergence between the views of our lay subjects and current law. In
the case of perjury, for example, lay participants were less likely than the law to
distinguish between lying in court under oath and lying to police while not
under oath, and between literally false statements and literally true but
misleading statements. Similarly, with respect to bribery, participants’ views
diverged significantly from current law. For example, respondents sought to
criminalize both commercial bribery and payments accepted by an office-holder
in return for performing a non-official act, despite the fact that neither form of
conduct is a crime under current American federal law.
PREVIOUS STUDIES OF COMMUNITY ATTITUDES REGARDING WHITE COLLAR
It is often said that those who commit white collar crimes are subject to less
severe punishments than those who commit street offenses.4 The usual
implication seems to be that such disparities are somehow unjust.5 But, on
reflection, it should be clear that treating a white collar crime less severely than
a street crime would be unjust only if the white collar crime in question was no
4. See, e.g., Ilene H. Nagel & John L. Hagan, The Sentencing of White-Collar Criminals in Federal
Courts: A Socio-Legal Exploration of Disparity, 80 MICH. L. REV. 1427 (1982).
5. E.g., JEFFREY REIMAN, THE RICH GET RICHER AND THE POOR GET PRISON: IDEOLOGY,
CLASS, AND CRIMINAL JUSTICE 126–29, 141–42 (8th ed. 2006).
No. 2 2012]
less blameworthy than the corresponding street crime. The problem is assessing
the relative blameworthiness of white collar crime.
A handful of recent studies have sought to measure the blameworthiness of
white collar crime in the eyes of the lay community. The data contained in this
literature can be divided into three general categories: (1) data concerning
perceptions of the relative seriousness of white collar crime compared to
“street” crime; (2) data concerning perceptions of the relative seriousness of
otherwise unrelated forms of white collar crime; and (3) data concerning
perceptions of where the line should be drawn between white collar conduct
that is treated as a crime and related white collar conduct that is not, and how
the punishment for such related forms of white collar crime should be graded.
One of the more carefully controlled studies of the first type was conducted
by Kristy Holtfreter and colleagues.6 Subjects were asked, “[w]ho do you think
should be punished more severely,” a person who commits a “street crime like
robbery and steals $1,000,” or a person who commits a “white-collar crime like
fraud and steals $1,000?” Subjects could also respond that the two crimes
should “receive equal punishment.”7 The study reported that “65.4 percent of
the sample felt violent offenders should receive harsher punishments.”8
Unfortunately, the Holtfreter study has several significant limitations. One is
that it seems to treat “fraud” as a proxy for white collar crime generally.9
Although fraud is certainly an important white collar offense, it is only one of a
collection of offenses typically designated as such, and no explanation is given
for why attitudes about fraud seriousness should be regarded as an adequate
stand-in for attitudes regarding white collar crime more broadly. A second
limitation is that the study did not specify exactly what occurred during each
offense. For example, robbery implies the use of force or violence and is often
perpetrated by means of a weapon. It is therefore possible that subjects
compounded the offense. It seems obvious, in any event, that robbery involves
more serious wrongs and more serious harms than theft by means of deceit.10
Thus, it is not surprising that a white collar offense, with no potential for
physical harm, would attract lesser condemnation than a potentially violent blue
collar offense. Unfortunately, that does not tell us as much as we would like to
know about the distinction between white collar and blue collar crime per se;
there are too many other relevant factors.11
The literature on community attitudes toward white collar crime also
contains data concerning how people rate the comparative seriousness of
essentially unrelated white collar crime behavior. The National Public Survey
on White Collar Crime, conducted by the National White Collar Crime Center
(NWCCC) in 2000 and again in 2005, is representative here. The study found,
for example, that respondents rated an insurance agent’s fraud as more serious
than a corporation’s reporting false quarterly earnings to increase the value of
its stock, but less serious than a pharmaceutical company’s releasing a new drug
while hiding information revealing important health and safety issues for
consumers.12 Such findings, while interesting and relevant to policymakers in
deciding how to enforce and punish various white collar crimes, do not directly
address the kinds of concerns identified above regarding the distinction
between closely related forms of white collar conduct.
Only a handful of studies address the issue considered here—namely, where
to draw the line between closely related white collar criminal and non-criminal
acts. One is a study of community attitudes regarding “cartel behavior”
conducted at the University of Melbourne Law School by Caron Beaton-Wells
11. In 2000, the National White Collar Crime Center took a similar, but less controlled, approach
when it asked participants to compare the seriousness of twelve crimes, including a street crime that
caused or threatened injury and two white collar crimes that caused injury. NAT’L WHITE COLLAR
CRIME CTR., THE NATIONAL PUBLIC SURVEY ON WHITE COLLAR CRIME (2000). Comparing armed
robbery with selling tainted meat, forty-five percent of respondents said that selling tainted meat was
more serious, thirty-six percent said that armed robbery was more serious, and nineteen percent said
that they were equal in seriousness. Id. at 12. Comparing armed robbery with failing to recall a
defective vehicle, forty-eight percent said that armed robbery was more serious, thirty-eight percent
said that failing to recall a defective vehicle was more serious, and thirteen percent said that they were
equal in seriousness. Id.
In 2005, the NWCCC conducted another study using a different set of twelve scenarios, some
of which were presumed by the authors to be white collar crimes. NAT’L WHITE COLLAR CRIME CTR.,
THE 2005 NATIONAL PUBLIC SURVEY ON WHITE COLLAR CRIME (2005). The scenarios included: (1)
a corporation falsely reports its quarterly earnings, (2) a pharmaceutical company hides safety
information about a new drug, (3) an insurance agent misrepresents coverage and grossly inflates costs,
(4) a group of hackers steals patient information from a hospital and then sells it, and (5) a physician
files false claims with an insurance company. One representative finding displays the limitations of this
data: It was found that a bank teller’s embezzling $10,000 from a customer was more serious than a
robbery of $100, and less serious than an offender’s attacking another person during a bar room fight
and causing serious injury. Here we have three offenses, one presumably white collar, two not. The
white collar offense falls between the two non-white collar offenses in terms of perceived seriousness.
The three offenses, however, have virtually nothing in common. It is therefore inappropriate to draw
any broad conclusions about the seriousness of white collar crime in general from comparing
embezzlement to either of the other two offenses. For another study following a similar approach, see
Nicole Leeper Piquero, Stephanie Carmichael & Alex R. Piquero, Assessing the Perceived Seriousness
of White-Collar and Street Crimes, 54 CRIME & DELINQUENCY 291 (2008).
12. THE 2005 NATIONAL PUBLIC SURVEY ON WHITE COLLAR CRIME, supra note 11, at 15.
No. 2 2012]
and colleagues.13 The Melbourne study focuses on three key forms of antitrust
misconduct: price fixing, market allocation, and output restrictions. Subjects
were asked to rank the seriousness of these three forms of conduct in
comparison to each other, to others forms of white collar-type conduct, and to
other forms of crime.14 The 2000 NWCCC study also addresses this sort of
comparison between closely related white collar conduct. Respondents were
asked to compare the seriousness of a bribe accepted by a public official to a
bribe offered by a private citizen—an issue further considered below.15 In our
view, the Melbourne and NWCCC studies reflect a useful approach to
measuring community attitudes on several difficult issues in white collar crime.
However, we think the studies would have been more useful if the scenarios
had been pitched at a lower level of generality.
EMPIRICAL STUDIES OF ISSUES IN WHITE COLLAR CRIME
To examine the distinctions between criminal and non-criminal activities
within three domains of white collar offenses (bribery and gratuities; perjury
and false statements; and fraud), we conducted four surveys. For each domain,
we outline the legal standards by which the conduct is evaluated, describe
potentially liminal cases, and present our empirical assessment of the views of
lay participants toward these cases. This analysis aims to compare and contrast
lay views and current legal standards.
Respondents were recruited from Amazon’s Mechanical Turk service,
which allowed for a diverse sample of American adults. Each study had a target
sample of fifty respondents. Data were discarded if the respondent had an
abnormally fast completion time (less than half of the median) or incorrectly
answered a question intended to screen inattentive participants.16 Each study
13. Caron Beaton-Wells, Fiona Haines, Christine Parker & Chris Platania-Phung, Survey of Public
Opinion on Cartel Criminalisation in Australia: Report of Preliminary Findings (2001),
14. Among the white collar offenses included were an insurance company denies a valid claim to
save money; a company director uses his position dishonestly to gain personal advantage; a company
misleads consumers about the safety of goods; a company fails to ensure worker safety; a company
evades government income taxes; a person uses inside information in deciding to buy or sell shares. Id.
15. THE NATIONAL PUBLIC SURVEY ON WHITE COLLAR CRIME, supra note 11, at 7. Seventy-four
percent of respondents said it was more serious for a public official to accept a bribe, 12% said it was
more serious for a private citizen to give a bribe, and 14% said they were equally serious. As indicated
supra text accompanying note 11, our findings on this issue were similar.
16. Cf. Daniel Oppenheimer, Tom Meyvis & Nicolas Davidenko, Instructional Manipulation
Checks: Detecting Satisficing to Increase Statistical Power, 45 J. EXPERIMENTAL SOC. PSYCHOL. 867
(2009). A question early in the study asked participants to rate three movies on a deliberately
nonsensical scale. If participants read the directions immediately above the question, they would know
to bypass it. If participants did answer the question, however, they were prompted to reread the
directions and check their response. Those who did not then correct their response were marked as
inattentive. Oppenheimer and colleagues show that those who miss such checks fail to notice subtle
distinctions in survey materials whereas those who pass the check (either initially or after being
prompted) are sensitive to such variations.
began with a brief description of its procedure. Participants were told that the
study evaluated how people act in various social situations. After giving their
consent, participants were asked a question that instructed them to bypass,
rather than answer it. Those who recorded an answer were marked as
inattentive. Then, respondents completed a page of individual difference
Following the individual difference measures, participants were given
instructions describing the format of the scenarios. Participants were told they
would view a core “story” with multiple possible “endings” and that it was for
them to determine which distinctions, if any, were relevant. After each scenario,
participants were asked three questions. First, they were asked to rate the moral
blameworthiness of the described act on a scale ranging from 1-Not at All
Blameworthy to 7-Very Blameworthy. Second, they were asked whether the act
should be treated as criminal (Yes or No). Third, they were asked how severely,
if at all, the person should be punished on a scale ranging from 1-No
Punishment to 7-Severely Punished.
Basic demographics (age, sex, occupation, educational attainment, state of
residence) were collected at the end of the study. Participants were also asked a
series of questions related to the set of scenarios presented in the study. For
example, in the bribery study they were asked whether they had ever run for
public office, held a position of responsibility at a larger firm, worked at a
company with a gifts policy, been involved in lobbying, or given money to a
A. Studies on Bribery and Gratuities
Since ancient times, virtually all systems of criminal law have criminalized
bribery.18 Bribery corrupts institutions by inviting inappropriate grounds for
decisionmaking. It creates political instability, distorts markets, undermines
legitimacy, retards development, and leads to injustice, unfairness, and
inefficiency.19 Traditionally, the offense has required a government official to
accept a cash payment in return for agreeing to perform some official act.20
Variations on the prototypical case raise numerous questions about the public’s
attitudes towards bribery.
First, in some of the most infamous cases of bribery, a briber is caught on
tape offering or giving a bribee a briefcase full of cash in return for some official
No. 2 2012]
act. Although this is arguably the paradigmatic case, modern statutes typically
define bribery more broadly to include offering “anything of value” to a public
official with the intent of influencing an official act. The term “thing of value”
has been read broadly to refer to a range of tangible and intangible things, such
as offers of future employment, unsecured short-term (and subsequently
repaid) loans, restaurant meals, tickets for athletic events, ostensibly valuable
(but actually worthless) stock certificates, and sexual favors.21 Under current
federal law, a defendant who accepts something of value other than cash or
tangible property would be subject to the same potential punishment as
someone who accepts cash or tangible property.22 One issue, then, is whether
the public would agree that the acceptance of other kinds of things of value—
such as services, political endorsements, and contributions to a political
campaign—should be regarded as equivalent to the acceptance of cash or
Second, bribery law has traditionally made it a crime to give something of
value for the purpose of influencing an “official act.” Therefore, payments given
to influence an unofficial act presumably do not constitute a bribe. But courts
have disagreed about what constitutes an “official act.” Some courts have
broadly construed the term to include acts that were not within the defendant’s
official duties.23 Others have read the statute narrowly and excluded acts that
were not sufficiently specific and pending.24 This raises the issue of the extent to
which the lay public would regard as wrongful the payment of money to
influence what we take to be an unofficial act, such as the giving of a political
Third, there may be a difference between soliciting or accepting a bribe, on
the one hand, and offering or giving of a bribe, on the other. As Green has
previously argued, soliciting or accepting a bribe involves a kind of disloyalty.25
Public officials are supposed to work in the best interests of their constituents or
institutions, rather than in the interests of third parties who tempt them.
Offering or giving a bribe, by contrast, involves a different dynamic: The briber
(in contrast to the bribee) normally does not have a duty of loyalty. Rather, he
induces another person to be disloyal. As such, he acts as an accomplice by
influencing, soliciting, inciting, or persuading another to do wrong. Other things
being equal, it is arguably less wrongful to induce another to do a wrongful act
21. GREEN, supra note 1, at 199.
22. See, e.g., United States v. Williams, 705 F.2d 603 (2d Cir. 1983) (interpreting 18 U.S.C. §
23. E.g., United States v. Parker, 133 F.3d 322 (5th Cir. 1998) (holding Social Security
Administration employee’s use of computer to fraudulently create documents was official act though
outside her official duties); United States v. Jefferson, 562 F. Supp. 2d 687 (E.D. Va. 2008) (finding
congressman’s meeting with foreign and U.S. government officials to promote interests of companies
with business interests in Africa constituted “official acts” within meaning of bribery statute).
24. E.g., Valdes v. United States, 475 F.3d 1319 (D.C. Cir. 2007) (finding police officer’s use of
computer to search police database was not official act because it did not involve any matter then
pending before the police department).
25. GREEN, supra note 1, at 203.
than to do the wrongful act oneself.26 For purposes of punishment, current law
does not distinguish between taking and giving a bribe. Would the lay public
nevertheless recognize such a distinction?
Fourth, as traditionally understood, bribery consists of a bilateral agreement
or quid pro quo in which the bribee solicits or accepts something of value from
the briber in exchange for the bribee’s acting, or agreeing to act, on the briber’s
behalf. This exchange requires a meeting of the minds, with the bribee agreeing
to “be influenced” in the performance of an official act. Sometimes, however,
people give money and other things to public officials not in return for
something specific, but merely as a gift or gratuity. Sometimes, the gift-giver is
trying to obtain goodwill. Other times, she is genuinely trying to say “thank
you” to the official for an unsolicited favorable act (and perhaps to buy
continued goodwill). Federal law distinguishes between giving or offering
something of value “for the purpose of influencing” an official act and doing so
“for or because of any official act.”27 The first act involves a quid pro quo and
constitutes true bribery. Under federal law, it can result in a sentence of up to
fifteen years in prison. The second act is a mere gratuity. The convicted
defendant faces a maximum penalty of only two years. One issue, then, is
whether the lay public accepts the distinction made by the federal statute.
Finally, bribery has traditionally involved the offer or payment of money to
a “public official,” defined under the leading federal statute as a member of
Congress, or official or employee of any branch of the federal government.28
The statute also applies to jurors, witnesses, and non-government employees
who occupy a position of trust with federal responsibilities, as in the case of an
employee of a private nonprofit organization that administered a sub-grant of a
municipality’s federal block grant.29 Yet there is a noticeable trend toward
criminalizing bribery in the commercial context. For example, federal law now
criminalizes bribes accepted by investment advisors, contestants in television
game shows, bank employees, sellers of alcoholic beverages, labor union
officials, railroad employees, and radio disc jockeys.30 Several states have
bribery statutes that apply to commercial contexts more generally.31 The
recently passed United Kingdom Bribery Act makes bribery a crime not only
for government officials, but also for persons in business.32 This raises the issue
26. Id. at 209.
27. 18 U.S.C. § 201(c)(1)(A) (2006).
28. Id. § 201(a)(1).
29. Id. § 201(b)(3), (4); Dixson v. United States, 465 U.S. 482, 496 (1984).
30. Id. § 215; id. § 205(c); 29 U.S.C. § 186 (2006); 49 U.S.C. § 11907 (2006).
31. For example, Texas law makes it a felony for a fiduciary, such as an agent, employee, trustee,
guardian, lawyer, physician, officer, director, partner, or manager to accept or agree to accept any
benefit from another person “on agreement or understanding that the benefit will influence the
conduct of the fiduciary in relation to the affairs of his beneficiary.” TEX. PENAL CODE ANN. § 32.43
32. U.K. Bribery Act, 2010, c. 23, § 3(2)(b) (Eng.).
No. 2 2012]
PERCEPTIONS OF WHITE COLLAR CRIME CULPABILITY 41
of the extent to which the public regards a bribe paid to a private employee in a
commercial context as comparable to a bribe paid to a government official in
the public context.
The above-discussed legal distinctions provide a large number of
briberyrelated questions to consider. Therefore, we address bribery in two separate
data collections. In the first, we address the first and second issues identified
above. In the second, we address issues three through five.
1. Bribery and Gratuities Study #1: Methods
The first set of scenarios concerned two issues: the wrongfulness of a
bribee’s accepting a direct payment of cash or other tangible property
compared to the bribee’s accepting a non-property-related thing of value or
receiving a thing of value only indirectly; and the blameworthiness of accepting
things of value in exchange for official acts versus accepting things of value in
exchange for unofficial acts.
Given that the leading American statute makes it a crime for a public
official to accept “anything of value” with intent to be influenced in an official
act,33 we wanted to know how respondents’ culpability determinations would
change if we varied the nature of the thing of value. Thus, in one scenario, the
putative briber, a company chief executive officer (CEO) named Reeves, wants
the putative bribee, a state legislator named Smith, to vote against legislation
that would hurt Reeves’ firm. In one variation, Reeves offered Smith $20,000
cash; in a second, Reeves offered $20,000 worth of renovations to Smith’s home;
in a third, Reeves offered an (otherwise legal) $20,000 campaign contribution;
and in a fourth, Reeves promised an endorsement in the next campaign. We
predicted that subjects would view the payment of cash as most wrongful, with
the in-kind home renovations a close second. A campaign contribution seemed
more likely to fall outside the scope of bribery. The American political system,
for better or worse, runs on campaign contributions, and although such
contributions are not technically supposed to be given in return for specific
promises, we anticipated that many subjects would not distinguish between
campaign contributions given with and without strings. Additionally, we
predicted that subjects would assume that the campaign contribution would not
personally benefit the would-be bribee, but would instead be used to run his
campaign. We thought that the promised endorsement would not be viewed as
bribery, but rather as merely “politics as usual,” an example of “you scratch my
back, and I’ll scratch yours.”
As described earlier, only money given in return for the performance of an
“official act” counts as bribery under federal law.34 We wanted to know the
extent, if any, to which our subjects would also think it was wrongful to give
money in return for the performance of an unofficial act. To this end, we
formulated a set of scenarios in which Smith, still a member of the state
33. 18 U.S.C. § 201 (2006).
34. Id. § 201(a)(3).
legislature, is approached by Johnson, a CEO who wants to run for mayor.
Johnson seeks Smith’s endorsement, which we assume is not an official act
(though we should say that we have not found any law directly on point). In
return for Smith’s endorsement, Johnson offers the same list of inducements as
in the previous scenario: cash, in-kind home renovations, a campaign
contribution, and a mutual political endorsement. We predicted that, other
things being equal, our subjects would rank the payment in return for an
unofficial act as considerably less wrongful than payment for an official act.
Fifty participants were recruited for this study from Amazon’s Mechanical
Turk service. Data from one participant were discarded due to an abnormally
fast completion time (less than half the median).35 Of the remaining forty-nine
participants (twenty male, twenty-nine female), the median age was thirty-six.
Fifty-three percent of participants had college degrees. The procedures
followed were as described above. The official and unofficial act stories were
presented on separate pages in counterbalanced order. The scenarios and their
various endings were not labeled.
2. Bribery and Gratuities Study #1: Results
Table 1 summarizes the study results, including the mean scores for the
blameworthiness and punishment measures and the percentage of participants
indicating that each act should be a criminal offense. Blameworthiness and
punishment were both higher when the offer was made for an official act than
for an unofficial one.36 A high percentage of respondents (approximately ninety
percent) said that offering money for a vote on legislation (an official act)
should be treated as a crime. A similarly high percentage said that offering
money for a political endorsement (which we regarded as an unofficial act)
should be treated as a crime. This was surprising and seemingly inconsistent
with current law, which does not treat payment for what we assume to be a
nonofficial act, such as a political endorsement, as a bribe.
35. Cf. Oppenheimer et al., supra note 16, at 867.
36. The data on blameworthiness and punishment severity were analyzed using a
repeatedmeasures ANOVA, with a two (official vs. unofficial act) by four (various means) design. The main
effect of official versus unofficial act was significant for both blameworthiness F(
) = 32.18, p < .001
and punishment F(
) = 60.09, p < .001.
No. 2 2012]
PERCEPTIONS OF WHITE COLLAR CRIME CULPABILITY 43
Agreeing to Give
Note: For each type of act (official vs. unofficial), scores for a given dependent
variable that do not share subscripts are significantly different from each other.
Standard deviations are in parentheses. Blameworthiness and punishment scores are
on scales ranging from one to seven.
The percentage of the sample criminalizing an activity and the
blameworthiness and punishment ratings varied as a function of the means by
which the act was solicited. Judgments of criminality followed the same pattern,
so they will be discussed together.37 When the bribee accepted an offer of cash,
over ninety percent of our subjects said that the act should be treated as a
crime, regardless of whether it was an official or (presumably) unofficial act.
When an offer of home renovations was accepted, a similarly large percentage
said that it should be a crime to agree to perform an official act; significantly
fewer (about seventy-eight percent) said that it should be a crime when the
agreement was for an endorsement.38 Over seventy percent said that accepting a
campaign contribution in return for performing an official act should be treated
as a crime, but only slightly more than forty percent said that accepting a
campaign contribution in return for giving an endorsement should be treated as
37. The effect of bribery means was also significant for both blameworthiness F(2.2, 106) = 60.98, p
< .001 and punishment F(2.1, 102) = 91.70, p < .001. Due to a sphericity violation, the Greenhouse–
Geisser correction was used for the relevant blameworthiness and punishment analyses.
38. Responses to the yes or no question asking whether the act should be treated as a crime were
analyzed using a repeated measures chi-squared in all studies.
a crime (perhaps because both acts involved an explicitly political process).
When the bribee accepted the offer of an endorsement, thirty-five percent said
that it should be a crime if an official act was requested, and less than ten
percent said it should be a crime if an unofficial act was requested.
We also investigated the role of individual difference factors in shaping
bribery attitudes. We created separate composite variables for the criminality,
blameworthiness, and punishment ratings for the official and unofficial acts. For
example, we averaged the four blameworthiness scores for each of the public
act cases to create an overall public blameworthiness score. In the case of the
criminality data, we scored a scenario marked as a crime as a one and a scenario
marked as no crime as a zero. Thus, the criminality averages reflect the
proportion of scenarios marked as crimes.
We then probed the relationship between individual difference measures
and each of the six composites (three for each act type) in a two-step process.
First, we examined the simple bi-variate correlations between the composites
and measures of educational attainment, sex, competitive world beliefs, political
orientation, and faith in various institutions. The individual difference items
that correlated with any of the dependent measures were then used as
predictors in a multiple regression. Predictors that were not significant in the
initial regression were removed from the final regression model.39 As was also
seen in subsequent studies, there were remarkably few significant effects. In this
study, the only variable that predicted views of either set of scenarios was
whether the participant had ever worked at a company that had a gift policy. If
the participant had, then they were more likely to criminalize (β = .42, p < .01),
and more severely blame (β = .34, p < .05), and punish (β = .32, p < .05)
attempts to influence the unofficial act. There was no such effect on attempts to
influence public actors. That the presence of a gift policy affected judgments for
the unofficial act but not the official act may in part be due to the relative lack
of ambivalence in responses to the official act. The overwhelming majority of
the sample was highly critical of attempts to influence a vote on a bill, which
sharply limited the potential role of individual differences.
Previous research has sometimes found that differences in punitiveness
toward white collar offenders depended on the gender, race, or educational
attainment of the participant.40 In these studies, we did not find differences
based on gender or educational attainment. Given our comparatively small
sample size, we cannot assume that such differences do not exist, rather only
that, if they do exist, the differences are likely small. The samples were not large
39. The use of multiple regression analysis was intended to deal with the conceptual overlap of
some of the individual difference measures. For example, in the fraud study, both social and economic
conservatism correlated with ratings of criminality, but only economic conservatism was a significant
predictor in the multiple regression; social conservatism only affected judgments of criminality to the
extent that it reflected economic conservatism.
40. For a useful summary, see Holtfreter et al., supra note 6, at 52; see also James D. Unnever,
Michael L. Benson & Francis T. Cullen, Public Support for Getting Tough on Corporate Crime: Racial
and Political Divides, 45 J. RESEARCH IN CRIME & DELINQUENCY 163 (2008).
No. 2 2012]
PERCEPTIONS OF WHITE COLLAR CRIME CULPABILITY 45
enough to test meaningfully for ethnic differences.
3. Bribery and Gratuities Study #2: Methods
The second set of scenarios concerned three issues: the wrongfulness of
accepting money or other things of value as a quid pro quo versus accepting
money or other things of value as a mere gift; the wrongfulness of a bribe
accepted by a public official versus a bribe accepted by an official of a private
company; and the wrongfulness of soliciting or accepting a bribe versus offering
or giving a bribe.
Federal law draws a significant distinction between giving or offering
something of value “for the purpose of influencing” an official act and giving or
offering something of value “for or because of any official act.”41 The first act
involves a quid pro quo and constitutes true bribery; it can result in a sentence
of up to fifteen years in prison. The second act is a mere gratuity with a
maximum penalty of two years. We wanted to see if our subjects would make a
similar distinction. We devised a set of scenarios in which the gift-giver (CEO
Larson) either has or has not had contact with the gift-recipient (state legislator
Jones) prior to the recipient’s official act (a vote on the location of a new
government office building). The scenario with prior contact describes a direct
bribe: money is offered in advance as a quid pro quo. In the scenario without
prior contact, the gift-giver gives a “thank you” gift of $20,000 after the vote.
The gift giver did not have a direct expectation of future services. We tested two
variations of the gratuity scenario. In one, the recipient has announced that he
will soon be retiring from office, which largely removes the possibility of future
inappropriate influence from the gratuity. In the other variation, the recipient
has no plans to retire and the gift giver is said to hope that his gift will “help his
company maintain good relations” with the recipient. We predicted that our
subjects would rate accepting a gratuity as less blameworthy than accepting a
bribe, and would rate the case in which the gift recipient would remain in office
as more blameworthy than that in which the recipient was about to retire.
Although bribery has traditionally involved the acceptance of money by a
“public official,” there has been a trend toward criminalizing bribery in the
commercial context. To determine how our subjects would view the
wrongfulness of commercial bribery, we formulated a set of parallel scenarios in
which CEO Larson is interacting with Heller, a board member of another
company. Heller is voting on the location of a new office building, but this time,
it is a private office building and the government is not involved. There were
three variants: a bribe, a gratuity to the retiring employee, and a gratuity to the
employee with no plans of retiring.
In all of these scenarios, the focus was on the culpability of the person
receiving the bribe or gratuity. But most statutes also make it a crime to offer or
give a bribe or gratuity, and, for purposes of punishment, do not distinguish
between the two acts. Thus, we posed an additional set of questions that asked
41. 18 U.S.C. § 201(b), (c) (2006).
subjects to assess the wrongfulness of Larson’s offering the bribe to the state
legislator. We expected most subjects to view accepting a bribe as more
wrongful than giving a bribe since only the former typically involves a breach of
loyalty; the latter presumably involves the aiding and abetting of another in his
The methodology for the second bribery study paralleled the first study.
Fifty-two participants were recruited from Amazon’s Mechanical Turk service.
Data from three were discarded due to abnormally fast completion times (less
than half the median) or incorrectly answering a question intended to screen
inattentive participants. Of the remaining forty-nine participants (fifteen male,
thirty-four female), the median age was thirty-three. Fifty-one percent of
participants had college degrees.
Participants were told they would view a core “story” with multiple possible
“endings,” and that they should determine which distinctions, if any, were
relevant. The stories about the government official (Legislator Jones) and
corporate employee (Heller) were presented on separate pages in
counterbalanced order. The questions about the wrongfulness of offering the
bribe to the state legislator were always last.
4. Bribery and Gratuities Study #2: Results
Table 2 summarizes the results of the study. The government official was
judged to be more blameworthy and deserving of more punishment than the
corporate employee.42 The distinction between bribes in the government
context and bribes in the commercial context was less significant than expected.
When the bribe was accepted by a board member of a large company, nearly
eighty percent of our subjects said that this should be treated as a crime.
Although this was slightly less than in the context of the bribe accepted by a
public official, eighty percent is still substantial. This was a striking finding given
that federal law does not make such conduct a crime at all, at least not under a
bribery statue.43 Attitudes of the American lay public were more consistent with
United Kingdom law, which, under the Bribery Act of 2010, treats commercial
bribery as indistinguishable from bribery in the public realm.44 It is also worth
noting that a significant number of respondents (about thirty-five percent)
believed that gratuities given in the commercial realm should be regarded as
42. The data on blameworthiness and punishment severity were again analyzed using a
repeatedmeasures ANOVA, this time with a two (government official versus corporate employee) by three
(various means) design. The case of the briber was set aside for this analysis. Blameworthiness F(
= 9.38, p < .01 and punishment F(
) = 21.18, p < .001 varied as a function of whether the official
worked for the government or a private corporation. Blameworthiness F(1.5, 73) = 50.14, p < .001 and
punishment F(1, 72) = 78.36, p < .001 also varied depending on whether the exchange between giver
and taker constituted a bribe or merely a gratuity. Greenhouse–Geisser was used in these analyses.
43. If a private official makes a decision that runs against the best interests of his company in
return for taking money for his own benefit, he may have breached a fiduciary duty to the company’s
shareholders, and thus be potentially liable for a different offense.
44. U.K. Bribery Act, 2010, c. 23 (Eng.).
No. 2 2012]
PERCEPTIONS OF WHITE COLLAR CRIME CULPABILITY 47
All else being equal, respondents always rated the acceptance of a quid pro
quo payment as significantly more blameworthy than the acceptance of a gift.
Indeed, the difference between a quid pro quo and a gift was greater than the
difference between a quid pro quo accepted by a government official and one
accepted by a corporate employee. Distinguishing between quid pro quo
payments and gifts is consistent with current federal law.
Finally, there were no significant differences between how respondents
viewed the blameworthiness of bribers compared to bribees. This is consistent
with current U.S. law, but runs counter to our hypothesis: We expected the
bribee to be judged more harshly than the briber because the bribee is the one
betraying a trust whereas the briber is merely inducing the bribee to do so. Such
judgments might have been counter-balanced, however, by the fact that the
briber showed “initiative” that the bribee arguably did not.
To analyze individual differences for this study, we created separate
composites for the private and public bribee cases and combined the bribery
and gratuity cases for each. Two individual difference factors were significant.
When the receiver of funds was a private employee, a subject who had worked
at a company with a gift policy was more likely to criminalize the private
official’s actions (β = .27, p = .05), rate them as blameworthy (β = .36, p < .05),
and assign harsher punishments (β = .37, p < .05). Subjects with a high score in
competitive world beliefs were somewhat less likely to criminalize a private
official accepting money (β = .27, p = .05), but competitive world beliefs did not
have an effect on the other measures. When the receiver of funds was a public
official, the same two factors were relevant. Subjects who had worked at a
company with a gift policy were more likely to criminalize the official’s actions
(β = .32, p < .05) and rate them as highly blameworthy (β = .34, p < .05).
Participants high in competitive world beliefs were again less likely to
criminalize the official’s actions (β = .39, p < .01) or rate them as highly
blameworthy (β = .39, p < .01). Gift policies and competitiveness did not
significantly affect preferred punishment for the public official.
B. Study on Perjury and False Statements
Under U.S. federal and analogous state law, two main offenses involve lying
in the context of government procedures. The first, perjury, consists of willfully
making a false material statement while under oath, typically in a judicial or
legislative proceeding.45 Case law makes it clear that in order to qualify as
perjury, a statement must be literally false rather than merely misleading.46
Under the federal perjury statute, the punishment is a maximum of five years in
prison. The second offense, false statements, consists of making a false
statement in a matter within the jurisdiction of a government agency, including
in the context of a law enforcement investigation.47 The defendant need not be
under oath. Although the Supreme Court has never ruled on the issue, a
majority of lower courts have held that, as in the case of perjury, a false
statement must be literally false;48 a minority of lower courts have held to the
contrary.49 Under the leading federal statute, 18 U.S.C. § 1001, the punishment
for false statements is also a maximum of five years (unless it involves terrorism,
in which case the maximum sentence is eight years).50 We wanted to know how,
other things being equal, the public would regard the wrongfulness of lying in
court under oath compared with lying to law enforcement officials while not
under oath. In addition, we wanted to know the extent to which our subjects
would, if all else were equal, distinguish between literally false statements and
literally true but misleading statements.
No. 2 2012]
PERCEPTIONS OF WHITE COLLAR CRIME CULPABILITY 49
We were also interested in the effect of the perjury or false statements:
Specifically, should it matter whether a witness’s lie leads to a third party being
falsely inculpated (and therefore unjustly convicted) or, conversely, to a third
party being falsely exculpated (and therefore unjustly acquitted)? Our legal
culture traditionally views wrongful conviction as worse than wrongful
acquittal.51 In the past, federal law treated certain kinds of falsely exculpatory
statements as exempt from criminal prosecution.52 But, under current federal
law, false statements that have the effect of unjustly exculpating are treated no
less harshly than false statements that have the effect of unjustly inculpating.
We wanted to see whether lay subjects would recognize the inculpation–
To test the three key questions concerning perjury and false statements, we
formulated a range of related scenarios. In all of them, a man named Walt
witnessed an automobile theft near his home. At the time of the theft, Walt
would normally have been at work. His company has a strict attendance policy
and he risked being terminated were it known that he had been late on the day
in question. Walt was therefore motivated to conceal his whereabouts on the
day of the crime and not cooperate with authorities. In all cases, the main
suspect in the crime was Dave, a man Walt did not know.
We wished to explore the distinction between making a literally false
statement and making a literally true (though misleading) statement.
Accordingly, in one variation, Walt was called as a witness by the prosecution in
a criminal trial, and offered (exculpatory) testimony that was literally false: he
said he was not at home and did not witness the crime. In actuality, he did see
Dave commit it. In the other variation, Walt offered testimony that, though
misleading, was literally true: he said that he normally left for work by the time
the theft occurred. The prosecutor took that to mean that he did not witness the
theft. In each case, we stated that Dave was acquitted despite his guilt. We
predicted that, other things being equal, subjects would rate the literally false
statement as more wrongful than the merely misleading one. Green has
previously characterized this as the principle of caveat auditor: In certain
circumstances, a listener is responsible or partly responsible for ascertaining a
statement’s truth before believing it. This principle applies to cases of merely
misleading statements but not to lying.53 When A lies to B, A is telling B that A
believes what A is saying. If A is mistaken about her own assertion or is lying,
51. See Alexander Volokh, N Guilty Men, 146 U. PA. L. REV. 173 (1997). The Hebrew Bible says
that a witness who falsely inculpates is to receive the same punishment as that which would have been
given to the falsely accused; no such similar punishment is applicable for statements that are falsely
exculpating. Deuteronomy 19:18–19.
52. See Brogan v. United States, 522 U.S. 398 (1998) (repudiating judicially created “exculpatory
no” doctrine, under which a statement that would otherwise violate 18 U.S.C. § 1001 was exempt from
prosecution if it conveyed false information in a situation in which a truthful reply would have
incriminated the interrogee, and if it was limited to simple words of denial rather than more elaborate
53. See GREEN, supra note 1, at 78.
LAW AND CONTEMPORARY PROBLEMS
then A is wholly responsible for B’s false belief. But if A merely misleads B
without making an assertion, A has not told B that A believes what A is saying
(since what A is saying is neither true nor false). Thus, it is partially B’s fault for
relying on an assertion that A has not made. The underlying idea, as explained
by Jonathan Adler, is “that each individual is a rational, autonomous being and
so fully responsible for the inferences he draws, just as he is for his acts. It is
deception, but not lies, that require mistaken inferences and so the hearer’s
responsibility.”54 We also wanted to know how the subjects would regard the
wrongfulness of lying in court under oath compared with lying to the police
while not under oath. As a baseline, we used the aforementioned scenario in
which Walt is called as a witness by the prosecution. Walt was explicitly
described as swearing to tell the truth and then lying. We also created a
variation in which the police questioned Walt at the initial stage of the
investigation. He again lies, but this time he is not under oath. Because of his
lie, the police do not pursue an investigation of Dave; the guilty man is free of
suspicion. We predicted that all else being equal, subjects would rate lying
under oath in a court proceeding as more wrongful than lying to police when
not under oath. We surmised that violating a sworn oath would add to the
wrongfulness of the act.
Finally, we explored the difference between testimony that is falsely
exculpatory and that which is falsely inculpatory. In all of the previous
scenarios, Walt’s various statements had the effect of helping to unjustly acquit
or divert suspicion from a guilty man. An additional variant placed Walt in a
courtroom under oath, but this time, he made a literally false statement. The
statement had the effect of allowing an innocent man to be convicted (he
testified that he did not witness the crime when, in fact, he saw someone other
than Dave commit it). We predicted that false statements that were falsely
inculpatory would be rated as significantly more wrongful than false statements
that were exculpatory. We thought that our subjects’ judgments would be
informed by the common maxim that “it is better that [five, ten, twenty, or a
hundred] guilty men go free than that one innocent man be convicted.”
1. Perjury and False Statements Study: Methods
The method for the perjury and false statements study closely paralleled the
method for the prior studies. Fifty-one participants were recruited from
Amazon’s Mechanical Turk service. Data from two were discarded due to
abnormally fast completion times (less than the median) or incorrectly
answering a question intended to screen inattentive participants. Of the
remaining forty-nine (twenty male, twenty-nine female), the median age was
thirty-three. Forty-nine percent of participants had college degrees.
Participants were told they would view a core “story” with multiple possible
“endings” and that they should determine which distinctions, if any, were
54. Jonathan E. Adler, Lying, Deceiving, or Falsely Implicating, 94 J. PHILOSOPHY 435, 437 (1997).
No. 2 2012]
PERCEPTIONS OF WHITE COLLAR CRIME CULPABILITY 51
relevant. The demographics page for this study also asked participants whether
they had ever testified in court or been questioned by the police.
2. Perjury and False Statements Study: Results
Table 3 summarizes the results and shows that blameworthiness scores were
high for all versions of the scenario. In all of the versions containing a direct lie,
Walt was rated within one interval of the top of the scale. Thus, while some lies
may have been seen as more blameworthy than others, all were viewed as
sufficiently wrong to reach the ceiling of the scale, thereby obscuring potential
differences between the lies. The version in which Walt merely misleads the
court, however, was seen as less blameworthy than all of the others (including
the parallel case in which he lied).55
Blameworthiness Punishment CriPmeirncaelnizting
Lie to Court, Exculpatory 6.45a (1.08) 4.12b (1.39) 93.9%a
Lie to Court, Inculpatory 6.47a (1.08) 4.63a (1.58) 89.8%ab
Lie to Police, Exculpatory 6.14a (1.15) 3.80b (1.73) 77.6%b
Mislead Court, Exculpatory 5.39b (1.71) 3.08c (1.68) 59.2%c
Note: For each dependent variable, scores that do not share subscripts are
significantly different from each other. Standard deviations are in parentheses.
Blameworthiness and punishment scores are on scales ranging from one to seven.
Punishment scores varied substantially more than blameworthiness scores.
As expected, a lie to a court that had an unjustly inculpatory effect was seen as
deserving greater punishment than a lie that had an unjustly exculpatory effect.
Additionally, a misleading statement with an exculpatory effect was seen as
deserving less punishment than a lie with an exculpatory effect. Contrary to
predictions, there was not a significant difference between a lie to a court with
an exculpatory effect and a lie to the police with the same exculpatory effect.
A majority of subjects thought that lying to police when not under oath
should be treated as a crime, though less than those who wanted to see lying
under oath treated as a crime. We think the distinction reflects the fact that, in
the case of perjury, the defendant has sworn to tell the truth, and the perception
that, other things being equal, lying in court might pose a greater potential for
harm than lying to the police. Inasmuch as subjects regarded perjury as the
more serious crime, their views were inconsistent with federal law, which
55. The data on blameworthiness and punishment severity were analyzed using a repeated
measures ANOVA. Scores varied across condition for both blameworthiness F(2, 116) = 14.79, p < .001
and punishment F(3, 144) = 17.58, p < .001. Greenhouse–Geisser correction was used in the
LAW AND CONTEMPORARY PROBLEMS [Vol. 75:33
regards false statements and perjury as equivalent in seriousness (both offenses
have a maximum of penalty of five years in prison, except when false statements
involves terrorism, in which case the maximum penalty is eight years).
The percentage of subjects who wanted to criminalize the conduct in each
version of the scenario was revealing. More respondents wished to criminalize
literally false statements to the court than wished to criminalize literally true
though misleading statements to the court. This corresponds to the moral
distinction between lying and merely misleading. Nevertheless, a majority of
subjects wished to criminalize even misleading statements made under oath.
This finding suggests a divergence between community attitudes and current
We examined the individual difference data in this study at the overall level
by creating a single set of composites across all four scenarios, and at the
individual level by looking at each scenario separately. Individual difference
measures did not significantly relate to criminality, blameworthiness, or
punishment judgments in either approach.
C. Study on Fraud
The term “fraud” has a wide range of meanings in Anglo-American criminal
law. At its core, fraud consists of the misappropriation of money or property by
means of deceit. But fraud is also often defined more broadly to include (1) acts
aimed at objects other than the misappropriation of property (such as the
deprivation of honest services, obstruction of governmental functions, and
obtaining of unjust advantage), and (2) acts committed by means other than
deception (such as breach of trust, conflicts of interest, and exploitation).56 As a
result, the concept of fraud is both ubiquitous and elusive.
We focused this study on what we took to be the core sense of fraud: theft
by deception. Even within this core conception, the precise boundaries of fraud
can be hard to pin down. Our free market system tends to respect and reward
aggressive business practices. What constitutes true fraud can be difficult to
distinguish from “sharp dealing,” “puffing,” or “seller’s talk.” The question,
then, is how to distinguish between cases of misrepresentation that constitute
criminal fraud and cases that do not.
The law has traditionally required that, to be fraudulent, a misleading
statement or lie must be material, in that it concerns the price, quantity,
effectiveness, or quality of the goods or services in question. The fraud must go
to the nature of the bargain itself, rather than to the circumstances surrounding
the bargain. For example, in a leading Second Circuit case, the defendants were
in the business of selling stationery supplies through salesmen who solicited
orders by telephone.57 In order to “get past” the secretaries who answered the
56. Though the recent decision in Skilling v. United States, 130 S. Ct. 2896 (2010), has limited the
scope of federal mail fraud, in holding that the honest services fraud criminalized by 18 U.S.C. § 1346
applies only to bribery or kickbacks and not to other fraudulent deprivations of intangible rights.
57. United States v. Regent Office Supply, 421 F.2d 1174 (2d. Cir. 1970).
No. 2 2012]
PERCEPTIONS OF WHITE COLLAR CRIME CULPABILITY
phone and speak to the purchasing agents directly, the salesmen would make
false representations by saying that they had been “referred” to the customer
by a mutual friend or another customer, or that they were “stuck” with
stationery ordered by a customer who had died. While not condoning the use of
such misrepresentations as a matter of business ethics, the Second Circuit held
that they did not constitute criminal fraud, since, crucially, there was no
evidence that any of the misrepresentations had been “directed to the quality,
adequacy or price of goods to be sold, or otherwise to the nature of the
We wanted to see whether the public would agree that the line between
criminal fraud and unethical business practices should be drawn where the
Second Circuit and other courts have placed it.59 We also wanted to know
whether the distinction between material and non-material misrepresentations
was sufficiently fine-grained. In particular, we were interested in seeing if our
subjects would regard certain misrepresentations regarding the quality,
adequacy, or price of goods as more serious than others.
Another issue in the law of fraud concerns the distinction between deceptive
affirmative statements and deceptive omissions. A fraud can be premised on
either, but a material omission is fraudulent only when there is a duty to
disclose or when the omission has the effect of making an affirmative statement
misleading by implication.60 Once an omission has been determined to be
fraudulent, however, it is treated the same as a deceptive affirmative statement.
We wanted to know the extent to which the public would regard various kinds
of omission as fraudulent and how they thought omissions should be punished.
Finally, under current law, the usual maximum penalty for mail fraud is
twenty years in prison.61 Theoretically, such penalties could apply regardless of
the value of the property stolen, but in practice, the value of the property
obtained is relevant under Sentencing Guidelines.62 We investigated the extent
to which, all else being equal, the value of property involved would affect
judgments about the appropriate punishment.
We created two parallel sets of scenarios involving potentially fraudulent
statements made in connection with the sale of a good. One set involved the
sale of an $800 laptop computer. The other set involved the sale of a $40,000
automobile. We predicted that, other things being equal, subjects would rank
fraud committed in connection with the more expensive item as more
blameworthy than fraud committed in connection with the less expensive item.
58. Id. at 1179.
59. United States v. Starr, 816 F.2d 94, 98 (2d Cir. 1989); United States v. Dinome, 86 F.3d 277,
284–85 (2d Cir. 1998).
60. See, e.g., Starkman v. Marathon Oil Co., 772 F.2d 231 (6th Cir. 1985), cert. denied, 106 S. Ct.
61. 18 U.S.C. § 1341 (2006). If the fraud occurs in the context of a national emergency or affects a
financial institution, the maximum penalty is thirty years.
62. U.S. SENTENCING GUIDELINES MANUAL § 2B1.1 (2011).
LAW AND CONTEMPORARY PROBLEMS [Vol. 75:33
We expected scenarios involving misrepresentations of quality to be judged
as most deserving of punishment. We distinguished between defects that
involved the particular good being sold and defects that involved the model
more generally. Thus, in one variation, the salesperson tells his would-be
customer that the product (whether the laptop or automobile) is a new, brand
name model, when it is really a refurbished model made with off-brand parts. In
another variation, the salesperson knowingly sells a product with a defective
part. In a third, the salesperson is directly asked whether the model (whether
the laptop or automobile) is known to have a specific problem: a tendency to
overheat (in the case of the laptop) or a tendency to burn oil (in the case of the
car). The salesperson assures the customer that it does not, even though the
salesperson knows that the model suffers from that problem. In a variation of
this last scenario, we had a case in which the salesperson is not asked about the
problem with the model and merely omits mentioning it. Here, we predicted
that our subjects would rate the omission as deserving less punishment than the
corresponding affirmative misstatements.
We predicted that scenarios involving two variants of “seller’s talk” would
be evaluated as substantially less serious. In one, the salesperson tells the
customer that the actual price of the good in question is higher than the special
sale price that he is offering to the customer ($1,000 versus $800 in the case of
the computer; $45,000 versus $40,000 in the case of the car), when in fact, the
lower price is the normal price. In the other, the salesperson tells his customer
that the product in question is “very popular among important business people
in New York City,” even though he knows this is not true. Both of these
statements are direct lies, but neither distorts the quality of the good in
question; they are not relevant to the fundamental bargain. We expected that
subjects would not believe them to be worthy of criminalization.
1. Fraud Study: Methods
The methodology for the fraud study closely paralleled the methodology for
the prior studies. Fifty participants (sixteen male, thirty-four female) were
recruited from Amazon’s Mechanical Turk service. No participants had
abnormally fast completion times (less than half the median) or incorrectly
answered a question intended to screen inattentive participants. The median
age was forty-one. Sixty percent of participants had college degrees.
Participants were told that they would be asked to rate the conduct of a
salesperson in various cases and that they should decide whether the differences
between the cases were important to their evaluations. The car and laptop
scenarios were presented on separate pages, with accompanying questions, in
counterbalanced order. The scenarios were not labeled.
At the end of the study, participants were also asked whether they had
worked in sales, had ever bought a laptop or computer after talking to a
salesperson, or considered themselves to be very knowledgeable about cars or
No. 2 2012]
PERCEPTIONS OF WHITE COLLAR CRIME CULPABILITY 55
2. Fraud Study: Results
The twelve fraud scenarios were divided into two groups (car versus laptop)
of six cases each (the various sales tactics). The primary data analysis on the
blameworthiness and punishment data took the form of a two by six repeated
measures ANOVA. The type of good did not have an effect on either measure,
so the more expensive good did not lead to perceptions that the sales tactics
were more blameworthy or deserving of more punishment. There was, however,
an interaction between sales tactic and type of good on the punishment
measure. When the salesperson falsely claimed the product was new, there was
elevated punishment in the car case relative to the computer case (see Table
4).63 For the other sales tactics, there were not differences between the
computer and car cases. Notably, falsely claiming that the product was new was
the only sales tactic that an overwhelming majority of participants rated as
criminal, perhaps explaining why it alone shows this distinction. Overshadowing
this interaction, however, was an effect of sales tactic on both blameworthiness
and punishment.64 Since these effects were highly consistent across both
measures and goods, we discuss them together.
63. F(5, 245) = 4.76, p < .001
64. Blameworthiness F(5, 245) = 48.43, p < .001; punishment F(5, 245) = 57.21, p < .001.
LAW AND CONTEMPORARY PROBLEMS [Vol. 75:33
The scenario judged to be the most worthy of criminalization (slightly less
than ninety percent in the case of the car and eighty percent in the case of the
laptop) involved a false claim that the particular item being sold was new and
brand name. This finding is consistent with the notion that true fraud involves
misrepresentations that fundamentally impact the nature of the bargain.
There was a significant drop-off from the scenario judged most worthy of
criminalization to the scenario judged next most worthy of criminalization:
when the salesperson explicitly lied about a problem with the model. Fifty-eight
percent of participants rated this as deserving of criminalization. We expected
this number to be higher because this misrepresentation also affects the nature
of the bargain. One possible explanation is that a general problem with the
model, rather than a problem with the particular unit, is something that the
No. 2 2012]
PERCEPTIONS OF WHITE COLLAR CRIME CULPABILITY 57
customer could have independently investigated and verified. Some participants
may have been holding the victim partially responsible in this case.
Less serious than the deliberate lie was selling a product with a known
defective component (between forty-two and forty-four percent of participants
criminalized). Again, this case is criminal fraud under law. But in this case, only
a minority of respondents rated it as such. One explanation is that the parts in
these cases (batteries for both products) were viewed as comparatively
insignificant. Alternately, our subjects might have made an implicit assumption
that such immediate failures would be covered by a warranty.
Our subjects made a clear distinction between affirmative statements and
omissions. When the salesperson affirmatively lied about a problem with the
models, close to sixty percent said it should be treated as a crime. When the
salesperson merely failed to mention the very same problem, less than thirty
percent said it should be treated as a crime.
The category rated least deserving of criminal sanctions included
misrepresentations thought to consist of non-material “seller’s talk.” When the
salesperson falsely claimed the product was “on sale,” only twenty percent of
respondents said the statement should be treated as a crime. When the
salesperson falsely claimed the product was “popular with business people in
New York City,” less than ten percent of respondents said this should be
treated as a crime. The difference between the two cases can likely be explained
by the fact that the first statement at least involved a question of the product’s
“price,” while the second was “puffing” in its purest form. To the extent that
subjects distinguished between misrepresentations concerning material facts
and misrepresentations concerning non-material facts, their judgment was
consistent with current law. But our subjects did not reliably criminalize all
material misrepresentations. In these cases, though, it is hard to say what
unspoken assumptions may have guided subjects.
The relatively high assessments of moral blameworthiness are of interest.
Although subjects seemed reluctant to brand these potentially fraudulent
activities as criminal, they did register their moral aversion. While these sales
strategies may be viewed as “lawful,” they were clearly not entirely acceptable
from subjects’ ethical points of view.
We again assessed the role of individual difference factors in shaping
criminal justice attitudes. In this case, we formed separate blameworthiness,
punishment, and criminality composites for the car and computer scenarios. For
both sets of scenarios, economic conservatism was the only significant predictor
of views on criminality (βcomputer = -.55, p < .001; βcar = -.42, p < .01) and
punishment (βcomputer = -.39, p < .01; βcar = -.39, p < .01). The more economically
conservative a person was, the less likely he or she was to judge a given act as
fraudulent and the less punishment imposed. Economic conservatism was
measured on a seven-point scale. For each step on the scale, from liberal to
conservative, a participant was 9.8% less likely to judge a computer scenario as
a case of fraud and 7.5% less likely to judge a car scenario as a case of fraud.
LAW AND CONTEMPORARY PROBLEMS [Vol. 75:33
Those respondents who viewed themselves as more conservative may have
taken a laissez faire attitude that markets are self-regulating and government
regulation, whether criminal or civil, is disfavored. No other measure had an
White collar criminal offenses, such as bribery, gratuities, perjury, false
statements, and fraud, reflect a kind of moral ambiguity that is lacking in the
case of many more familiar street crimes. Whether a given actor will be
punished as a criminal, or admired as a successful businessperson, politician,
lobbyist, or witness will often depend on nuanced moral judgments and subtle
distinctions in facts. Because the ability of laypersons to distinguish between
criminality and non-criminality and to make confident judgments about
retributive desert is crucial to the functioning of the criminal justice system, we
conducted a set of studies to assess lay attitudes regarding a variety of issues in
white collar crime.
In general, laypersons in our studies were capable of making fairly
finegrained distinctions regarding white collar crime. In some cases, the distinctions
made by our respondents were consistent with current law, and therefore lend
weight to the view that the law draws distinctions in the appropriate places. For
example, participants in the fraud study were mostly sensitive to the kind of
distinctions that lie at the heart of fraud law, including the distinction between
misrepresentations relating to the quality, adequacy, or price of goods or
services, on the one hand, and mere “puffing,” on the other. The study, thereby,
produced a pattern of results largely consistent with prevailing doctrines.
In other cases, however, we found that the judgments made by our subjects
differed in some significant ways from current law. In the domain of perjury, for
example, public perceptions of seriousness diverged from current law,
particularly with respect to the distinction between lying in court under oath
and lying to police while not under oath. The familiar distinction between lying
and merely misleading was also less sharp than under current law. Similarly, in
the case of bribery, participants’ views diverged significantly from current
American federal law. Participants wanted to criminalize both commercial
bribery and payments accepted in return for performing a non-official act,
neither of which would normally be criminalized under the status quo. In these
latter cases where public attitudes diverged from current law, further
investigation is in order.
The moral intuitions of the lay public are an increasingly important
component in criminal law theory. In the context of white collar crime, a part of
criminal law that remains largely under-theorized and misunderstood, there is
an even greater need for this kind of research. These data present some
significant, if preliminary, insights into how the lay public would draw the
boundaries between criminal and non-criminal white collar behavior. These and
PERCEPTIONS OF WHITE COLLAR CRIME CULPABILITY 59
future studies should serve to aid theorists and policymakers seeking to sensibly
and coherently define the borders of legitimate business practice.
6. Kristy Holtfreter , Shanna Van Slyke, Jason Bratton & Marc Gertz , Public Perceptions of White-Collar Crime and Punishment , 36 J. CRIM . JUST. 50 ( 2008 ). For studies employing a similar approach , see Francis T. Cullen, Jennifer L. Hartman & Cheryl Lero Johnson, Bad Guys: Why the Public Supports Punishing White-Collar Offenders , 51 CRIME, L. & SOC. CHANGE 31 ( 2008 ); Francis T. Cullen, Richard A. Mathers, Gregory A. Clark & John B. Cullen , Public Support for Punishing WhiteCollar Crime: Blaming the Victim Revisited?, 11 J. CRIM . JUST. 481 ( 1983 ). Some of this literature also focuses on differences in community attitudes towards white collar crime based on demographic factors, such as race, gender, income, political orientation, and the like .
7. Holtfreter et al., supra note 6 , at 53.
8. Id . (emphasis removed).
9. Id. at 57. Though, to be fair, the study does suggest that “future research” should “includ[e] several different types of offenses .”
10. This point is discussed in a prior study conducted by the present authors, in which subjects were asked to compare the seriousness of twelve different means of committing a theft of a $350 bicycle . Green & Kugler, supra note 3 . The two forms of robbery (armed robbery and simple robbery) were ranked first and fourth in terms of seriousness, respectively. The two forms of fraud (passing a bad check and false pretenses) were ranked ninth and tenth in terms of seriousness, respectively .
17. C.G. Sibley & J. Duckitt , Big-Five Personality , Social Worldviews, and Ideological Attitudes: Further Tests of a Dual Process Cognitive-motivational Model, 149 J. SOC. PSYCHOL . 545 ( 2009 ). These included self-rated political orientation (1-Very Conservative to 7-Very Liberal), faith in various public institutions (government, courts, defense attorneys; 1-Not much faith to 7-A lot of faith), and eight items from the Competitive World Beliefs scale (a measure of support for dog-eat-dog social Darwinist beliefs ).
18. For a historical perspective, see generally JOHN T . NOONAN, BRIBES ( 1984 ).
19. GREEN, supra note 1, at 202.
20. See , e.g., 18 U.S.C. § 201 ( 2006 ) (leading American bribery statute ).
45. The federal statute is 18 U.S.C. § 1621 ( 2006 ).
46. Bronston v. United States , 409 U.S. 352 ( 1973 ).
47. 18 U.S.C. § 1001 ( 2006 ). Under federal law, the false statement is typically made to a government entity or official, such as an FBI, IRS, or SEC agent; under analogous state law, it is typically made to a state or local entity or official, such as a police officer .
48. United States v. Gahagan , 881 F.2d 1380 , 1383 ( 6th Cir . 1990 ); United States v . Vesaas , 586 F.2d 101 , 104 ( 8th Cir . 1978 ); United States v . Lozano, 511 F.2d 1 , 5 ( 7th Cir . 1975 ).
49. E.g., United States v . Stephenson , 895 F.2d 867 , 873 ( 2d Cir . 1990 ).
50. 18 U.S.C. § 1001 ( 2006 ).