CJPS4

Journal of Criminal Law and Criminology

Volume 81
Issue 3 Fall Article 6

Fall 1990

Rational Choice, Deterrence, and Social Learning
Theory in Criminology: The Path Not Taken

Ronald L. Akers

Follow this and additional works at: https://scholarlycommons.law.northwestern.edu/jclc

Part of the Criminal Law Commons, Criminology Commons, and the Criminology and Criminal
Justice Commons

This Criminology is brought to you for free and open access by Northwestern University School of Law Scholarly Commons. It has been accepted for
inclusion in Journal of Criminal Law and Criminology by an authorized editor of Northwestern University School of Law Scholarly Commons.

Recommended Citation
Ronald L. Akers,

Rational Choice, Deterrence, and Social Learning Theory in Criminology: The Path Not Taken

, 81 J. Crim. L. &
Criminology 653 (1990-1991)

https://scholarlycommons.law.northwestern.edu/jclc?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol81%2Fiss3%2F6&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarlycommons.law.northwestern.edu/jclc/vol81?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol81%2Fiss3%2F6&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarlycommons.law.northwestern.edu/jclc/vol81/iss3?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol81%2Fiss3%2F6&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarlycommons.law.northwestern.edu/jclc/vol81/iss3/6?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol81%2Fiss3%2F6&utm_medium=PDF&utm_campaign=PDFCoverPages

https://scholarlycommons.law.northwestern.edu/jclc?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol81%2Fiss3%2F6&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/912?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol81%2Fiss3%2F6&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/417?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol81%2Fiss3%2F6&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/367?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol81%2Fiss3%2F6&utm_medium=PDF&utm_campaign=PDFCoverPages

http://network.bepress.com/hgg/discipline/367?utm_source=scholarlycommons.law.northwestern.edu%2Fjclc%2Fvol81%2Fiss3%2F6&utm_medium=PDF&utm_campaign=PDFCoverPages

0091-4169/90/8103-653
THEJOURNAL OF CRIMINAL LAw & CRIMINOLOGY Vol. 81, No. 3
Copyright Q 1990 by Northwestern University, School of Law Printed in U.S.A.

Rational Choice, Deterrence, and Social
Learning Theory in Criminology: The Path

Not Taken*

Ronald L. Akers

I. INTRODUCTION AND OVERVIEW

“Rational choice” theory, which is derived mainly from the ex-
pected utility model in economics,’ has become a “hot” topic in
criminology, sociology, political science, and law. The evidence is
compelling: respectedjournals have published a major collection of
papers as well as several recent articles on the theory;2 sociological
treatises and articles have been published in the 1980s on both
macro- and microrational choice models;3 and finally, James S.
Coleman launched a new interdisciplinary journal, Rationality and So-
ciety, in 1989 with the (perhaps overblown) claim that “[w]ork based

* This article is a revision of a paper presented at the annual meeting of the
American Society of Criminology in Reno, Nevada, in November, 1989. Please direct
any inquiries or correspondence to Ronald L. Akers, Department of Sociology,
University of Florida, Gainesville, Florida 32611.

I See, e.g., J. HEINEKE, ECONOMIC MODELS OF CRIMINAL BEHAVIOR (2d ed. 1978)
[hereinafter J. HEINEKE; M. REYNOLDS, CRIME BY CHOICE: AN ECONOMIC ANALYSIS
(1985); Becker, Crime and Punishment: An Economic Approach, 76J. POL. ECON. 169 (1968).

2 THE REASONING CRIMINAL: RATIONAL CHOICE PERSPECTIVES ON OFFENDING (D.
Cornish & R. Clarke eds. 1986) [hereinafter THE REASONING CRIMINAL]; Williams &
Hawkins, The Meaning of Arrest for Wife Assault, 27 CRIMINOLOGY 721 (1989); Klepper &
Nagin, The Deterrent Effect of Perceived Certainty and Severity of Punishment Revisited, 27 CRIMI-
NOLOGY 721 (1989) [hereinafter Klepper & Nagin]; Paternoster, Decisions to Participate In
and Desist From Four Types of Common Delinquency: Deterrence and the Rational Choice Perspec-
tive, 23 LAw & Soc. REV. 7 (1989) [hereinafter Paternoster, Decisions to Participate]; Pater-
noster, Absolute and Restrictive Deterrence in a Panel of Youth: Explaining the Onset,
Persistence/Desistance, and Frequency of Delinquent Offending, 36 SoC. PROBS. 289 (1989)
[hereinafter Paternoster, Absolute and Restrictive Deterrence]; Piliavin, Thorton, Gartner &
Matsueda, Crime, Deterrence and Rational Choice, 51 AM. Soc. REV. 101 (1986) [hereinafter
Piliavin].

3 J. COLEMAN, INDIVIDUAL INTERESTS AND COLLECTIVE ACTION: SELECTED ESSAYS

(1986) [hereinafterJ. COLEMAN, INDIVIDUAL INTERESTS]; R. COOTER & T. ULEN, LAW AND
ECONOMICS (1987); M. HECHTER, PRINCIPLES OF GROUP SOLIDARITY (1987); R. POSNER,
ECONOMIC ANALYSIS OF LAW (3d ed. 1986); Donahue, Law and Economics: The Road Not
Taken, 22 LAw & Soc. REV. 903 (1988) [hereinafter Donahue]; Opp, What Can We Learn
From The Utilitarian Tradition?, 4 CONTEMP. SOC. 445 (1988) [hereinafter Opp]; Posner,
Comment on Donahue, 22 LAw & Soc. REV. 927 (1988).

RONALD L. AKERS

on a theory of rational action is the most rapidly developing of so-
cial theory and research.”‘ 4 Rational choice may be on the verge of
becoming for the 1980s and 1990s what neo-Marxist perspectives
were in the 1970s, as it spreads to virtually all social science disci-
plines and law.

The introduction of rational choice models into modern crimi-
nology as part of this general movement may prove to be valuable.
Much of the rational choice literature takes a strong quantitative
modeling approach derived from econometric modeling, which ad-
vances our ability to test complex models of criminal behavior and
the criminal justice system. Rational choice also has inspired some
empirical work on decision making in specific crime and crime
events as well as in criminal justice policy, both of which were
projects that might not otherwise have been done. This article
posits, however, that thus far, no new general theoretical concepts
or propositions have been added to criminological theory by ra-
tional choice studies.

A. DETERRENCE AND RATIONAL CHOICE

The utility premise of rational choice theory has an obvious af-
finity for the deterrence doctrine in criminology. Deterrence and
the utilitarian view of rational human nature have been with us since
at least the eighteenth century. The deterrence doctrine, which was
at the heart of classical criminology, arguably has been the most
researched topic in criminology since the latter part of the 1960s.5

Deterrence theory applies utilitarian philosophy to crime. “Ra-
tional choice” is based on economic theory derived from the same
utilitarian tradition. Both theories assume that human actions are
based on “rational” decisions-that is, they are informed by the
probable consequences of that action. According to the deterrence
theory, the rational calculus of the pain of legal punishment offsets
the motivation for the crime (presumed to be constant across of-
fenders but not across offenses), thereby deterring criminal activity.
In comparison, the rational choice theory posits that one takes those
actions, criminal or lawful, which maximize payoff and minimize
costs.

Despite the long historical connection suggested by their com-
mon utilitarian source, rational choice did not enter criminology
primarily as research or theory on deterrence; instead, it was first

4 SAGE Advertising Brochure (1989).
5 See, e.g., J. GIBBS, CRIME, PUNISHMENT AND DETERRENCE (1975); G. VOLD & T. BER-

NARD, THEORETICAL CRIMINOLOGY (3d ed. 1986).

654 [Vol. 81

SOCIAL LEARNING THEORY

introduced through economic analysis of crime.6 Later, criminolo-
gists involved in research on deterrence utilized the economic
model of rational choice to modify or expand the deterrence doc-
trine. The link between deterrence and rational choice has since be-
come well-established in the literature. 7

B. DETERRENCE AND RATIONAL CHOICE AS SPECIAL CASES OF SOCIAL

LEARNING PRINCIPLES

Neither deterrence nor rational choice theory is a general or
complete model of criminal behavior. The central concepts and
propositions in each-fear of legal punishment in deterrence theory
and the reward/cost balance (or expected utility function) in ra-
tional choice theory-are subsumable under the more general dif-
ferential reinforcement formula in social learning theory.
Differential reinforcement refers to the overall balance of rewards
and punishment for behavior. It encompasses a full range of behav-
ioral inhibitors and facilitators: rewards/costs; past, present, and
anticipated reinforcers and punishers; formal and informal sanc-
tions; legal and extra-legal penalties; direct and indirect punish-
ment; and positive and negative reinforcement, whether or not
rationally calculated.8 Some of the rational choice models of crime
in the literature have been expanded beyond the basic expected util-
ity proposition to include family and peer influences, moral judg-
ments, and other variables. 9 These, too, are reiterations of
concepts or variables derived from social learning theory,’ 0 and, to
some extent, from other extant criminological theories such as so-
cial bonding. I

This article’s principal thesis is that the primary concepts and
valid postulates of deterrence and rational choice are subsumable
under general social learning or behavioral principles. Further, I
will show that the relevance of learning principles for deterrence has

6 See R. CROUCH, HUMAN BEHAVIOR: AN ECONOMIC APPROACH (1979) [hereinafter R.
COUCH]; J. HEINEKE, supra note 1; Becker, supra note 1.

7 See supra note 2.
8 R. AKERS, DEVIANT BEHAVIOR: A SOCIAL LEARNING APPROACH (3d ed. 1985); Bur-

gess & Akers, A Diferential Association-Reinforcement Theory of Criminal Behavior, 14 Soc.
PROBS. 128 (1966) [hereinafter Burgess & Akers]; Akers, Krohn, Lanza-Kaduce &
Radosevich, Social Learning and Deviant Behavior: A Specific Test of a General Theory, 44 AM.
Soc. REV. 635 (1979) [hereinafter Akers & Krohn, Social Learning and Deviant Behavior].

9 See THE REASONING CRIMINAL, supra note 2; Paternoster, Decisions to Participate, supra
note 2; Paternoster, Absolute and Restrictive Deterrence, supra note 2.

10 See R. AKERS, DEVIANT BEHAVIOR: A SOCIAL LEARNING APPROACH (1973); Burgess
& Akers, supra note 8.

11 See T. HIRSCHI, THE CAUSES OF DELINQUENCY (1969).

1990]

655

RONALD L. AKERS

been largely overlooked, and their relevance to rational choice has
been almost completely missed in the criminological literature.

Accordingly, I agree with Homanst 2 and Opp, 13 both of whom
contend that rational choice is a special case of general behavioral
exchange or learning principles. Thus, leading theorists outside the
criminology field have recognized, to some degree, the connection
between rational choice and behavioral models.14 At the same time,
it is difficult to find any level of awareness in the rational choice
literature in criminology. The introduction of behaviorist theory
into criminology has resulted chiefly from the reformulation of

12 See G. HOMANS, CERTAINTIES AND DOUBTS: COLLECTED PAPERS, 1962-1985 (1987);

Homans, Collective Choice, 16 CONTEMP. Soc. 769 (1987) [hereinafter Homans, Collective
Choice].

13 See Opp, supra note 3.
14 Coleman initially referred to exchange theories as “broader social theories.” J.

COLEMAN, THE MATHEMATICS OF COLLECTIVE ACTION (1973) [hereinafter J. COLEMAN,
MATHEMATICS]. He subsequently recognized that social exchange and behavioral soci-
ologists, such as Homans, really pioneered the introduction of the basic principles appli-
cable to rational choice into sociology. SeeJ. COLEMAN, INDIVIDUAL INTERESTS, supra note
3. However, rather than drawing on the same behavioral principles that Homans did,
Coleman developed a rational choice model that advocates an abstract, mathematical
application of microeconomic analysis to non-economic behavior. See especially J. COLE-
MAN, INDIVIDUAL INTERESTS, supra note 3, at 85-136; see alsoJ. COLEMAN, MATHEMATICS,

this note, at 32-60. He explained, “[T]he direction I was pursuing lay in the paradigm of
micro-economic theory, with the abstract conception of rational economic man as the
central element in the theory.” Id. at 5. Further, “[i]t is theory which rests upon the
central postulate of economic theory, that of rational man attempting to pursue his self-
ish interests.” Id. at 15. Finally, he concluded, “[t]he social system is thus conceived as
a system of social exchange which functions like an economic market.” Id. at 93.

In contrast, Homans viewed his original work on the elementary forms of social
behavior as incorporating both “elementary psychology” and “elementary economics.”
See G. HOMANS, SOCIAL BEHAVIOR: ITS ELEMENTARY FORMS (1961). He still sees “behav-
ioral psychology as the more general theory from which ‘rational choice’ may itself be
derived.” Homans, Collective Choice, supra note 12, at 770. In his review of Coleman,
Homans states that while he supports rational choice theory, he also believes that it is
too limited; it is only a special case of behavioral principles. Id. Opp agrees with Ho-
mans that there is little difference between rational choice and behavioral theorists. See
Opp, supra note 3.

Economic theorists seem largely unaware of the fact that when their models are
applied outside of economics and thereby allow for less than pristine, purely rational
calculus, they become virtually indistinguishable from social exchange. “[R]ational
choice theory is deficient in its almost total neglect of developments in social psychol-
ogy.” Opp, supra note 3, at 446. The neglect is not total, however. See R. HOGARTH &
M. REDDER, RATIONAL CHOICE: THE CONTRAST BETWEEN ECONOMICS AND PSYCHOLOGY

(1987) (comparing and contrasting rational choice models in economics and psychol-
ogy). Heath, a British sociologist, also sees rational choice and social exchange theory
as the same approach, although he is critical of Homans and Blau and prefers to use the
former term. See A. HEATH, RATIONAL CHOICE AND SOCIAL EXCHANGE: A CRITIQUE OF
EXCHANGE THEORY (1976) [hereinafter A. HEATH]. For a critical, comprehensive review
of the applicability of behavioral principles in sociology and criminology, see J. GIBBS,
CONTROL: SOCIOLOGY’S CENTRAL NOTION (1989) [hereinafterJ. GIBBS, CONTROL] (trea-
tise on “control” as the central notion in social and behavioral sciences).

656 [Vol. 81

SOCIAL LEARNING THEORY

Sutherland’s differential association theory with behavioral princi-
ples (“differential association-reinforcement”) beginning in the
mid-1960s and continuing today.15 Behaviorist theory long has
been referred to as social learning theory, and it has been cited
widely and tested in the research literature. Indeed, a discussion of
social learning theory has become a standard feature in many crimi-
nology and delinquency textbooks. All of the theory’s central con-
cepts and propositions are accessible easily in the literature of the
past twenty years. It is in no sense an obscure or esoteric perspec-
tive in criminology. Nonetheless, the literature on rational choice in
crime has overlooked it.16 Even though some earlier deterrence re-
searchers had taken note of learning concepts, deterrence research-
ers today tend to skip over it when integrating deterrence and
rational choice theories.

II. DETERRENCE AND SOCIAL LEARNING

The relationship of deterrence concepts to social learning the-
ory is not a recent discovery. Akers, who saw the research on deter-
rence to be partial tests of certain aspects of social learning, clearly
proposed that a relationship existed many years ago. In 1977, he
wrote:

Research done on the deterrence of criminal behavior through legal
sanctions and threat of punishment is relevant and tends to be consis-
tent with social learning …. The deterrence research does not pro-
vide a full test of social learning, however, because the reinforcement
for criminal behavior that the threat of legal penalties must offset and
other reinforcement contingencies surrounding the behavior are not
measured in that research. 17

He continued in 1985:
Deterrence researchers have not referred to or presented their find-
ings as tests of social learning …. The importance of informal social
sanctions and normative definitions of the acts has been largely ig-
nored. Some of the more recent deterrence studies, however, take
note of social learning theory and report findings that are consistent
with the theory such as informal social and personal sanctions, moral

15 See AKERS, DEVIANT BEHAVIOR: A SOCIAL LEARNING APPROACH (1973); id. (3d ed.
1985); Burgess & Akers, supra note 8; Akers, Burgess &Johnson, Opiate Use, Addiction and
Relapse, 15 Soc. PROBs. 459 (1968) [hereinafter Akers, Opiate Use]; see also Jeffery, Criminal
Behavior and Learning Theory, 56J. CRIM. L. & CRIMINOLOGY 294 (1965).

16 Although I believe that I have located the major articles published on rational

choice and crime, it is a burgeoning literature, and there could be some analysis which
explicitly tries to show the connections between rational choice and social learning that I
have missed.

17 R. AKERS, DEVIANT BEHAVIOR: A SOCIAL LEARNING APPROACH 56 (2d ed. 1977).

1990] 657

RONALD L. AKERS

condemnation and normative commitment. 18

Many of the authors cited by Akers1 9 did not refer specifically to
social learning theory, however, and very few deterrence research-
ers in more recent years have commented on the relevance of social
learning. One exception to this generalization is Lanza-Kaduce,
who explicitly referred to social learning theory in his analysis of
perceptual deterrence and drunk driving.2 0 Most efforts to update
or revise the deterrence doctrine, however, make no reference what-
soever to social learning. 2′

Threat of legal punishment is one source or indicator of aver-
sive stimulus under the general concept of differential reinforce-
ment (balance of rewarding and aversive stimuli). Empirical tests of
social learning theory have long included measures of both “formal
deterrence” (perceived probability of getting caught by the police)
and “informal parental deterrence” (perceived probability of being
caught by parents). In both variables, the term “deterrence” is used
because the measures referred only to perception of the likelihood
of punishment. Not surprisingly, neither variable has much direct
effect because each refers only to variation in perceived likelihood of
aversive consequences. Researchers using the general concept of
differential reinforcement have included other variables which mea-
sure both rewarding and aversive consequences and the balance of
positive and negative reactions from peers and parents; these vari-
ables, in contrast, have strong effects. 22 Similar measures of infor-

18 R. AKERS, DEVIANT BEHAVIOR (3d ed.), supra note 8, at 54.
19 Akers cited the following authors on the importance of social and personal sanc-

tions. See Anderson, Chiricos & Waldo, Formal and Informal Sanctions: A Comparison of
Deterrent Effects, 25 Soc. PROB. 103 (1977) [hereinafter Anderson, Formal and Informal
Sanctions]; Grasmick & Green, Legal Punishment, Social Disapproval, and Internalization as
Inhibitors of Illegal Behavior, 71 J. CRIM. LAW & CRIMINOLOGY 325 (1980) [hereinafter
Grasmick & Green]; Paternoster, Saltzman, Waldo & Chiricos, Perceived Risk and Social
Control: Do Sanctions Really Deter?, 17 LAw & Soc. REV. 457 (1983) [hereinafter Paterson,
Perceived Risks and Social Control]. Akers also cited the following authors for the concepts
of moral condemnation and normative commitment. See Erickson, Gibbs & Jensen, The
Deterrence Doctrine and the Perceived Certainty of Legal Punishments, 42 AM. Soc. REV. 305
(1977);Jensen, Erickson & Gibbs, Perceived Risk of Punishment and Self-Reported Delinquency,
57 Soc. FORCES 57 (1978); Meier & Johnson, Deterrence as Social Control: The Legal and
Extralegal Production of Conformity, 42 AM. Soc. REV. 292 (1977); C. TITTLE, SANCTIONS
AND DEVIANCE (1980).

20 Lanza-Kaduce, Perceptual Deterrence and Drinking and Driving Among College Students,
26 CRIMINOLOGY 321 (1988).

21 See Klepper & Nagin, supra note 2; Miller & Anderson, Updating the Deterrence Doc-
trine, 77 J. CRIM. L. & CRIMINOLOGY 418 (1986); Piliavin, supra note 2; Williams & Haw-
kins, The Meaning of Arrest for Wife Assault, 27 CRIMINOLOGY 163 (1989); Paternoster,
Decisions to Participate, supra note 2; Paternoster, Absolute and Restrictive Deterrence, supra
note 2.

22 See Akers & Cochran, Adolescent Marijuana Use: A Test of Three Theories of Deviant

658 [Vol. 81

SOCIAL LEARNING THEORY

mal sanctions have been included in other research reports. 23

These findings underscore the fact that failure to locate a deter-
rent effect of the threat of legal punishment, while directly relevant
to the deterrence doctrine, says little by itself about general social
learning theory. Social learning subsumes deterrence, but only in
the context of the larger picture of-differential reinforcement. With-
out taking into account the other contingencies surrounding the be-
havior, fear of punishment by itself will be unrelated or weakly
related to criminal acts. Since these other measures include punish-
ing reactions (actual or perceived), they have sometimes been re-
ferred to as “informal deterrence.” However, they are more than
measures of informal deterrence; they are measures of an overall
balance of perceived costs and rewards. Formal deterrence is only
an incomplete indicator of differential reinforcement.

Social learning theory specifically incorporates both informal
social rewards and punishments as well as the “formal application of
sanctions by the legal and correctional system to control violation of
norms.” 24 It also includes “direct references to the specific deliber-
ate efforts of the formal control system to deter deviance [and] to
the deviance-preventing effects of formal and legal sanctions.” ’25

Akers further explains that:
‘Effective social control’ can easily be interpreted to mean that the so-
cial sanctions successfully reinforce conventional behavior and extin-
guish deviant behavior by rewarding conformity and punishing
nonconformity. The structure of social control in society arranges
contingencies of reinforcement in such a way that most people are
kept in line most of the time. Conforming behavior is successively
shaped over time (socialization) and becomes largely self-con-
trolled …. Moreover, we remain liable to at least intermittent rein-
forcement for conformity and punishment for deviance. . . . Each
person must also contend with the consequences (mainly punishment)
which the formal and legal control agents attach to his or her behavior,
but these consequences are often remote and uncertain and they
therefore have less impact than the person’s immediate primary
groups. 26

Behavior, 6 DEVIANT BEHAv. 323 (1985) [hereinafter Akers & Cochran, Adolescent Mari-
juana Use]; Akers & Krohn, Social Learning and Deviant Behavior, supra note 8.

23 See Krohn, Lanza-Kaduce & Akers, Community Context and Theories of Deviant Behavior:
An Examination of Social Learning and Social Bonding Theories, 25 Soc. Q. 353 (1984) [herein-
after Krohn, Community Context]; Krohn, Skinner, Massey & Akers, Social Learning Theory
and Adolescent Cigarette Smoking: A Longitudinal Study, 32 Soc. PROB. 455 (1985) [hereinaf-
ter Krohn, Social Learning Theory and Smoking]; Lanza-Kaduce, Cessation of Alcohol and Drug
Use Among Adolescents: A Social Learning Model, 5 DEVIANT BEHAV. 79 (1984).

24 R. AKERS, DEVIANT BEHAVIOR (3d ed.), supra note 8, at 34-35.
25 R. AKERS, DEVIANT BEHAVIOR (2d ed.), supra note 17, at 38.
26 Id. at 65.

1990] 659

RONALD L. AKERS

He continued:
The concept of refraining from deviance because of the cost in lost
stakes in conformity is one example of the more general concept of
negative punishment-one refrains from doing something not because
of the fear of direct punishment but because of actual or anticipated
loss of the reward or investment connected with alternative behavior.
[The concept of negative reinforcement refers to the other side of the
coin-engaging in acts as a way of avoiding aversive consequences.]
The differential reinforcement concept in social learning theory incor-
porates deterrence, reward-cost balance, positive and negative punish-
ment, positive and negative reinforcement, and other rational and non-
rational cognitive and behavioral processes of reward and punishment.2 7

Deterrence locates variation of criminal behavior in only one
part (direct positive punishment of criminal behavior) of one side of
the overall reinforcement equation, albeit including the three mo-
dalities of certainty, severity, and celerity. In its classical formula-
tion, deterrence really includes only one specific indicator of
positive punishment-namely, fear of legal penalties. As Gibbs
states:

[T]he proper definition [of deterrence] … is narrow. In a legal con-
text, the term ‘deterrence’ refers to any instance in which an individual
contemplates a criminal act but refrains entirely from or curtails the
commission of such an act because he or she perceives some risk of
legal punishment and fears the consequence.2 8

The full behavioral formula in social learning theory includes
both positive and negative punishment and positive and negative
reinforcement. It also includes schedules of reinforcement, imita-
tion, associations, normative definitions (attitudes and rationaliza-
tions), discriminative stimuli, and other variables in both criminal
and conforming behavior. When the deterrence doctrine is ex-
panded to encompass other variables beyond actual or perceived
risk of legal sanction, such as formal and informal social sanctions
and both rewards and punishment, it is no longer distinctively de-
terrence theory. It becomes something else: to some, it becomes
rational choice theory; I would counter that it simply moves even
closer to social learning theory.

27 Akers, A Social Behaviorist’s Perspective on Integration of Theories of Crime and Deviance, in
THEORETICAL INTEGRATION IN THE STUDY OF DEVIANCE AND CRIME: PROBLEMS AND PROS-
PEcTS 23, 31 (S. Messner, M. Krohn & A. Liska eds. 1989) [hereinafter Akers, A Social
Behaviorist’s Perspective] (emphasis added).

28 Gibbs, Punishment and Deterrence: Theory, Research and Penal Policy, in LAw AND THE
SOCIAL SCIENCES 319, 325-26 (L. Lipson & S. Wheeler eds. 1986) [hereinafter Gibbs,
Punishment and Deterrence].

660 [Vol. 81

SOCIAL LEARNING THEORY

III. SOCIAL LEARNING AND THE RATIONAL CHOICE MODEL OF

CRIME

Although the link between deterrence and rational choice has
been made in the literature, proponents of the rational choice per-
spective purport to offer much more than just an expansion of the
deterrence theory. They propose to offer a new, integrative per-
spective to all of criminology-theory, research, and policy. 29 How
credible are these claims? What does the “rational choice” model
bring to criminology? To answer these questions, we need to re-
view three main dimensions of the rational choice perspective as
they relate to current theory in criminology. These dimensions are
(a) the “rationality” of criminal acts and careers; (b) the actual or
perceived balance of rewards and costs associated with committing
crime or engaging in alternative behavior; and (c) the background
and other relevant variables.

A. THE ASSUMPTION OF A RATIONAL COMPONENT IN CRIMINAL

BEHAVIOR: DOES RATIONAL CHOICE THEORY DIFFER FROM

OTHER THEORIES?

The key issue is whether the rational choice perspective pro-
poses a purely “rational man” theory of criminal behavior. Does the
model argue for a direct resurrection of classical criminology or an
unmodified adoption of expected utility theory? Does it propose
that each person approaches the commission of a crime with a
highly rational calculus of pleasure and pain before acting? Does it
propose that the person chooses, with full free will and knowledge,
whether to commit a crime, taking into account only a carefully rea-
soned, objective or subjectively perceived set of costs and benefits?
Is the model essentially free of all constraining, positivistic, or deter-
ministic elements? If the answer to these questions is in the affirma-
tive, such that adherence to a strict rationality model of behavior is
truly the distinguishing feature of a rational choice model of crimi-
nal behavior, then it does offer something different to criminology.
It is also patently false.

As presented in the literature, however, this is not the rationality
assumption in actual rational choice models of crime. Regardless of
the assumptions upon which rational choice is based in the classical
economic model, the literature in criminology emphasizes limita-
tions and constraints on rationality through lack of information,
structural constraints, values, and other “non-rational” influences.a0

29 See THE REASONING CRIMINAL, supra note 2.
3o The notion of bounded rationality operating in voluntary action has been under-

1990] 661

RONALD L. AKERS

Indeed, the rational choice models in the literature go well beyond
this to paint a picture of partial rationality with various situational
and cognitive constraints and deterministic notions of causes and
motivations. In fact, some of the specific models are indistinguish-
able from current “etiological” or “positivistic” theories. Paradoxi-
cally, the assumption of a high level of rationality in behavior is not
crucial to current rational choice models of crime.

Coleman, who is the chief proponent of rational choice models
in sociology sees serious “deficiencies” in directly applying the eco-
nomic model of rationality to “real players,” even in very restricted
social situations.3 1 Donohue maintains that, as applied to law and
society, the assumptions of rationality and individualism are “im-
portant but not defining characteristics of economic methodology.”3 2

Heath believes that the pure rational choice model of riskless
choice, full knowledge, and no mistakes is untenable.33 He argues,
however, that these and other assumptions (such as the requirement
that information be collected prior to a decision and that the deci-
sion must be made slowly) are fallaciously attributed to rational
choice models and are not actually used by rational choice theorists.

Gibbs insists that the assumption of rationality and the question
of free-will versus determinism are not really relevant to the deter-
rence doctrine, or, I would add, to rational choice models of
crime.3 4 According to Gibbs, deterrence can be construed as either
free will or causation, without any difference in empirical predic-
tions. The notion of rationality is very vague, and categorical asser-
tions that all persons or all behavior are rational are indefensible.35

Gibbs later proposes that the rationality argument is irrelevant for
the general concept of social control and its efficacy. “[T]he notion
of rationality is intolerably vague, and an empirically applicable defi-
nition of rational behavior is bound to be arbitrary . . . if rational
behavior is defined as simply goal-oriented behavior.. . then virtu-
ally all of human behavior is rational and the rational-irrational dis-
tinction has no real consequences.- 36

Although the literature refers to “the reasoning criminal,” and

stood in sociology at least since Parson’s classic statement of his theory of social action
as socially structured voluntary choices in which the actor selects among socially struc-
tured means to achieve goals. See T. PARSONS & E. SHILS, TOWARD A GENERAL THEORY
OF ACTION (1951); see alsoJ. COLEMAN, MATHEMATICS, supra note 14, at 33-34.

31 J. COLEMAN, INDIVIDUAL INTERESTS, supra note 3.
32 Donahue, supra note 3, at 913 (emphasis added).
33 A. HEATH, supra note 14.
34 Gibbs, Punishment and Deterrence, supra note 28.
35 Id.
36 J. GIBBS, CONTROL, supra note 14, at 394.

662 [Vol. 81

SOCIAL LEARNING THEORY

the “rational component” in crime, there is great effort to show how
limited, circumscribed, and partial that rationality is. The most fre-
quently cited source on rational choice and crime is The Reasoning
Criminal, edited by Cornish and Clarke.37 Yet, Cornish and Clarke
do not propose a pure rational choice model. Rather, the assump-
tion in their model is simply that,

offenders seek to benefit themselves by their criminal behavior; that
this involves the making of decisions and choices, however rudimen-
tary on occasion these processes might be; and that these processes
exhibit a measure of rationality, albeit constrained by limits of time
and ability and the availability of relevant information …. 38

They see offenders as “reasoning decision makers” because
criminals “exercise some degree of planning and foresight and adapt
their behavior to take account of proximal and distal contingen-
cies.”‘ 39 The degree of planning does not have to be great to be
considered rational by Cornish and Clarke, and they warn that “one
must be wary of definitions of rationality that rely too much on evi-
dence of planning.”40

Thus, Cornish and Clarke assert a very minimal assumption of
rationality, which does not seem to differ very much from the level
of rationality assumed in most criminological theories. In fact, it re-
ally does not even differ much from the assertions by Katz in his
analysis of the “seductions” of crime, which he views as the very
antithesis of rational, utilitarian explanations of crime.41 The em-
pirical chapters in Cornish and Clarke provide support for limited,
rather than pure, rationality in crime. This is in spite of the fact that
Cornish and Clarke acknowledge that the empirical studies in the
volume are mainly limited to economic offenses, which they pre-
sume to be the most rational of crimes. While a chapter on violent
crimes is not included, a chapter on “opioid addiction” is.42 The
author of this chapter, Bennett, explains addiction as rational “deci-
sion-making” because he found some evidence that addicts were
able to control opiate use at times and finds little evidence of “com-
pulsion, irrationality. . . or mindlessness in the decision to take the
drug.”43 This test of rationality is minimal indeed; virtually all ex-
tant criminological theory would pass it.

37 THE REASONING CRIMINAL, supra note 2.
38 Id. at 1.
39 Id. at 13 (emphasis added).
40 Id. at 14.
41 J. KATz, SEDUCTIONS OF CRIME (1989).
42 Bennett, A Decision-Making Approach to Opioid Addiction, in THE REASONING CRIMI-

NAL, supra note 2, at 83 [hereinafter Bennett].
43 Id. at 90.

1990] 663

RONALD L. AKERS

Carroll and Weaver do not go as far as Bennett, but they do
make a distinction between the “strong form” of rationality in the
economists’ expected utility model (which they find is empirically
wrong) and the more valid “limited rationality” form which charac-
terizes rational choice models in criminology. 44 Similarly, Lattimore
and Witte describe the purely rational model of expected utility as
lacking in empirical support from studies of both criminal and non-
criminal behavior.45 According to Lattimore and Witte, the axioms
and assumptions of this strong rationality model do not hold due to
cognitive limitations, short-cut decisions, inconsistent value prefer-
ences, and other constraints on rational action. They propose a
“prospect theory model” based on limited rationality and utility
functions. Other authors also start with a classical economic model,
then go on to show the inappropriateness of that model for criminal
behavior. They find that rational maximization of expected utility in
the commission of crime is the exception rather than the rule and
propose models of crime with limited rationality.46

Hirschi sees control theory and rational choice theory as essen-
tially the same theory, but coming from different disciplinary back-
grounds. He bases this connection mainly on the rational choice
content in his social bonding theory. 47 In his original statement of
the theory, Hirschi explicitly identifies the bond of “commitment”
as rational; however he does not contend that this is a wholly ra-
tional calculation of consequences of deviance. 48 Indeed, he makes
it clear that there are severe restrictions on the exercise of rational-
ity.49 Paternoster 56 presents what he labels a “deterrence/rational
choice” model of delinquent behavior, in which he invokes Matza’s

44 See Carroll & Weaver, Shoplifters’ Perceptions of Crime Opportunities: A Process-Tracing
Study, in THE REASONING CRIMINAL, supra note 2, at 20 [hereinafter Carroll & Weaver].

45 See Lattimore & Witte, Models of Decision Making Under Uncertainty: The Criminal
Choice, in THE REASONING CRIMINAL, supra note 2, at 130 [hereinafter Lattimore & Witte].

46 See, e.g., Johnson & Payne, The Decision to Commit a Crime: An Information-Processing
Analysis, in THE REASONING CRIMINAL, supra note 2, at 170; Tuck & Riley, The Theory of
Reasoned Action: A Decision Theory of Crime, in THE REASONING CRIMINAL, supra note 2, at
156 [hereinafter Tuck & Riley]; Walsh, Victim Selection Procedures Among Economic Criminals:
The Rational Choice Perspective, in THE REASONING CRIMINAL, supra note 2, at 40.

47 Hirschi, On the Compatibility of Rational Choice and Social Control Theories of Crime, in
THE REASONING CRIMINAL, supra note 2, at 106 [hereinafter Hirschi, Compatibility].

48 Id. at 120-21.
49 While I think he is wrong in claiming that no other major theory has a “rational

choice content,” id. at 110- 11, Hirschi is justified in viewing rational choice and social
bonding as highly compatible. However, he goes beyond this to claim that they are
essentially the same theory, albeit coming from different disciplines.

Rational choice theory and social control theory share the same image of man, an
image rather different from the image of sociological positivism. Rational choice
theory and social control theory are therefore the same theory reared in different
disciplinary contexts.

664 [Vol. 81

SOCIAL LEARNING THEORY

“soft determinism” 5′ in order to stress that the model includes both
rational free will and determinism. Rational choice models, accord-
ing to Paternoster, recognize that there are “choice structuring”
variables, and that the choices do not involve complete information
or rational analytic methods. Although he states that the model is
meant to explain the “decision” to participate, to continue, or to
desist from delinquency, his measures of the dependent variable are
the usual ones of self-reported delinquent behavior. Therefore,
there is no way in his empirical model to separate the decision to act
from the delinquent act itself. There is nothing in his model to dis-
tinguish it from any other theory of delinquency regarding assump-
tions of rationality. In fact, as I shall show below, his model is
simply a combination of variables taken from social learning and so-
cial bonding theory.

It should be apparent from this review that rational choice the-
ory does not assume.that all or even most criminal acts result from
well-informed calculated choices. The rational choice models in the
literature leave room for all levels of rationality, except the most
mindless, pathological, and irrational. In this regard, therefore,
they do not differ from social learning, social bonding, or, contrary
to what Hirschi argues, other sociological and social psychological

Id. at 113.
If control and rational choice are the same theory, then there is no need to deal

with them separately. If social bonding and social learning are compatible, then both
are compatible with rational choice. See Akers, A Social Behaviorist’s Perspective, supra note
27. Rational choice and social bonding are not, however, the same theory. Social bond-
ing’s element of commitment incorporates the notion of costs of criminal behavior,
which is one variable at the heart of the rational choice theory. On the other hand,
rational choice explicitly hypothesizes deterrent effects from both legal and extra-legal
sanctions. Social bonding theory neither explicitly incorporates the deterrence doctrine
nor analyzes the effects of rewards/punishment on conformity and deviant behavior.

While it would be incorrect to say that control theory has no room in it for the
direct effects of sanctions, such effects are not explicitly included in the explanation of
conformity/deviance. Social bonding theory directly includes neither deterrence con-
cepts nor more general concepts of rewards/punishment through sanctions. Hirschi
does not discuss deterrence, and apparently does not consider it part of the commitment
or rational component in his theory. Similarly, although he does describe parental su-
pervision, he does not analyze sanctioning of behavior by parents or others. Specific
reference to direct parental and peer rewarding or punishing reactions to behavior is
absent from both Hirschi’s theory and empirical tests of his theory. Thus, the only way
in which social bonding theory conceptualizes the reward/cost balance is with reference
to costs through loss of investment in conformity (commitment). Control theory ig-
nores variation in the rational (or non-rational) inducements to delinquency against
which the cost is weighed.

50 See Paternoster, Absolute and Restrictive Deterrence, supra note 2; Paternoster, Decisions
to Participate, supra note 2.

51 D. MA-ZA, DELINQUENCY AND DRIFr (1964).

1990] 665

RONALD L. AKERS

models of crime.52

Yet does not social learning theory have a mechanistic view of
humans as automatons whose behavior is operantly conditioned and
shaped by forces of which they are no more aware than the rats,
pigeons, and monkeys on which operant conditioning principles
were first developed? Does not the social learning approach reject
mental constructs, cognition, and the reasoning human being? The
answer to both of these questions is “No.” Social learning theory
falls within the behaviorist tradition that makes “explicit theoretical
use of notions about cognitive and symbolic processes.” 53 Bandura
has gone beyond the cognitive processes incorporated into his ear-
lier version of social learning. 54 He developed a “social cognitive
theory” without eliminating the behavioral elements from it. He
views research on deterrence as providing support for his theory,
which includes concepts such as “symbolizing,” “forethought,” and
“self-reflective capabilities.” Although Bandura does not comment
directly on the issue, he apparently sees no contradiction between
his social behaviorism and the presumed rationality in deterrence
theory. 55

As applied by Akers to crime and deviance, social learning is a
behavioral approach to socialization which includes individuals’ re-
sponses to rewards and punishments in the current situation, the
learned patterns of responses they bring to that situation, and the
anticipated consequences of actions taken now and in the future in
the initiation, continuation, and cessation of those actions. It is a
“soft behaviorism” that allows for choice and cognitive processes. 56

It views the individual’s behavior as responding to and being condi-
tioned by environmental feedback and consequences. It does not
view the individual as unreasoning and only passively conditioned.
Although the theory proposes that conditioning can take place with-
out an awareness or knowledge of the connection between one’s be-
havior and its consequences, it also views the individual as
cognitively engaged, adapting to existing and anticipated conse-
quences, exercising self-reinforcement, and learning to respond to
environmental and internal cue stimuli.

While there is little difference in the minimal level of rationality

52 See Hirschi, Compatibility, supra note 47.
53 R. AKERS, DEVIANT BEHAVIOR (3d ed.), supra note 8.
54 See A. BANDURA, SOCIAL FOUNDATIONS OF THOUGHT AND ACTION: A SOCIAL COGNI-

TIVE THEORY (1986) [hereinafter A. BANDURA, SOCIAL FOUNDATIONS]; A. BANDURA, SO-
CIAL LEARNING THEORY (1977).

55 See A. BANDURA, SOCIAL FOUNDATIONS, supra note 54.
56 See R. AKERS, DEVIANT BEHAVIOR (3d ed.), supra note 8.

666 [Vol. 81

SOCIAL LEARNING THEORY

assumed by the theories, do rational choice models lean more to-
ward the rationality end of the continuum, while social learning and
other theories lean toward the non-rationality end? Perhaps, but if
proponents of rational choice theory maintain that their models
contain a larger element of rationality than can be found in social
learning or other current theories, then they need to do at least two
things. First, they must contrast careftilly the amount of rationality
assumed in their models with that assumed in the principal theories
in criminology. Second, they must show clearly how that difference
in assumed rationality increases our understanding of crime and de-
viance beyond the understanding gained from other theories.

Thus far, the rational choice theorists do neither of these. In-
deed, in the few places where “wider criminological theories” are
discussed, the emphasis is on the compatibility, not the differences
in assumptions, of rational choice with the other theories. Pains are
taken to show how limited the rational choice perspective is if it le-
ans too far toward pure rationality and does not take into account
the variables stipulated in these other theories. 57 No attempt is
made to demonstrate that there is a crucial difference in the assump-
tion of rationality or propositions about rationality in crime between
rational choice models and any other coherent theory, let alone to
demonstrate that the difference results in a superior explanation of
crime and deviance.

Instead, whenever a sharp contrast is drawn, it is likely to be
drawn between rational choice models and something vaguely re-
ferred to as “traditional criminology.” Traditional criminology 58 is
usually depicted in these comparisons as emphasizing “irrational”
or “pathological” elements in crime. 59 However, what constitutes

57 See Tuck & Riley, supra note 46.
58 THE REASONING CRIMINAL, supra note 2; R. CROUCH, supra note 6; Bennett, supra

note 42.
59 Our deeply held and abiding fears about crime depict it as irredeemably alien to
ordinary behavior-driven by abnormal motivations, irrational, purposeless, unpre-
dictable, potentially violent, and evil…. In the past, repeated attempts by criminol-
ogists to identify differences between criminal and noncriminal groups that could
explain offending have reinforced assumptions that offenders are similar to each
other and different from everybody else ….

… By contrast, the present volume.., develops an alternative approach, termed the
“rational choice perspective’ to the explanation of criminal behavior…. While it
does not deny the existence of irrational and pathological components in some
crimes, it suggests we examine more closely the rational and adaptive aspects of
offending.

Cornish & Clarke, Introduction to THE REASONING CRIMINAL, supra note 2, at v-vi (empha-
sis added).

Since Cornish and Clarke do not give references and specific theories are not dis-
cussed, we do not know either to which repeated past efforts and criminologists they
refer, or to which theories rational choice is offered as an alternative.

6671990]

RONALD L. AKERS

traditional criminology is not identified. If “traditional” means all
perspectives other than rational choice, then the portrayal is wrong
and the comparison meaningless. Aside from Freudian psychoana-
lytic theory which views virtually all deviant acts (and many con-
forming acts) as the outcome of irrational, unconscious motivation
or early biological explanations, which view the criminal acting from
uncontrollable genetic forces,60 what major theories of crime rely
on the irrational or pathological factors mentioned by Cornish and
Clarke as characterizing all of criminology? The answer is, “None.”

Similarly, Crouch contrasts the rational choice model with what
he thinks is prevailing criminological theory, namely one that views
the criminal as a:

[U]niquely motivated individual with a flawed character structure-
which drives him or her to engage in so-called deviant behavior, re-
gardless of the consequences. According to this view, criminal behav-
ior is unique, special, and deviant; hence there is an identifiable group
of persons possessed of a criminal mentality, which they will indulge,
impulsively or compulsively, without regard to the consequences of
their activities.6 1

This is obviously a highly distorted and uninformed appraisal of
criminological theory. Crouch made no effort whatsoever to com-
pare his economic model with either the real criminological theories
current at the time of his writing or the theories of earlier times. If
he had undertaken this comparison, he would have discovered that
the rejection in his model of the assumption that most criminals be-
have impulsively without regard to consequences does not distin-
guish it in the least from any other major contemporary
criminological theory.

We have known at least since Cressey showed us thirty-five
years ago that offenses labelled as “compulsive,” “totally sense-
less,” or “impulsive” can be understood with the same explanation
as other seemingly more rational crime.62 The categorization of
some act or series of acts as kleptomania, pyromania, or senseless
violence reflects the explanatory predisposition of the one making
the characterization (and the social characteristics of the offender)
more than it reflects the mental state of the offender. Building on
Cressey, Akers offered a social learning explanation of supposedly
compulsive, irrational crime, violent and nonviolent, in which such
behavior is depicted as not totally unplanned or unpredictable. Far

60 G. VOLD & T. BERNARD, THEORETICAL CRIMINOLOGY (3d ed. 1986);J. WILSON & R.

HERRSTEIN, CRIME AND HUMAN NATURE (1985).
61 See R. CROUCH, supra note 6, at 117.
62 Cressey, The Differential Association Theory and Compulsive Crimes, 45 J. CRIM. L. &

CRIMINOLOGY 29 (1954).

668 [Vol. 81

SOCIAL LEARNING THEORY

from being senseless, totally irrational, or compulsive, such behav-
ior is usually quite understandable with knowledge of the person’s
past learning, perceptions of current situations, and anticipation of
the consequences. If the person believes the acts are justified or
excusable and will not result in receiving more violence than given
(or alternatively that the acts are worth the risk in some other way)
then the more likely he or she will be to commit them.63 Therefore,
no new element of rationality is added to existing theory when Cor-
nish and Clarke argue that “even in the case of offenses that seemed
to be pathologically motivated or impulsively executed, it was felt
that rational components were also often present …. 64

B. REWARDS AND COSTS IN CRIMINAL BEHAVIOR

As previously noted, at the center of the rational choice model
is an objective (or subjective) utility proposition that the decision to
commit crime is a function of the balance of rewards and costs for
crime and its alternatives. 65 In this regard, rational choice theory
does propose something that is either absent or only implicit in bio-
logical, personality, strain, labelling, conflict, Marxist, or Suther-
land’s original differential association theories. It expands on the
deterrence doctrine, with which it is most often linked, and goes
beyond control theory with which it is increasingly linked.66 How-
ever, the explicit reference to rewards and punishments in rational
choice explanations of criminal behavior, does not take rational
choice theory beyond that which is already present in social learning
theory.

In social learning, instrumental or operant learning is the basic

63 R. AKERS, DEVIANT BEHAVIOR, supra note 10, at 209-15.
64 THE RATIONAL CRIMINAL, supra note 2, at 2.
65 In a sentence, the models of economic choice theory, of which the criminal
choice is a special case, hypothesizes that all individuals, criminal and non-criminal
alike, respond to incentives; and if the costs and benefits associated with an action
change, the agent’s choices are also likely to change.

J. HEINEKE, supra note 1, at 2. Similarly,
[t]his approach assumes that the individual contemplating a criminal act will decide
to commit the crime only if he or she expects that committing the crime will lead to
a more satisfactory outcome than not doing so.

Lattimore & Witte, supra note 45, at 130. Finally,
[t]he central assumption of a rational choice model of offending is that persons
make conscious decisions to offend based upon information about offenses, deci-
sions whose outcomes they believe will be beneficial or profitable to them.

Paternoster, Decisions to Participate, supra note 2, at 10.
66 See Hawkins & Williams, Acts, Arrests and the Deterrence Process: The Case of

Wife Assault (Aug. 1989) (unpublished paper presented to the Society for the Study of
Social Problems, Berkley, CA); Hirschi, Compatibility, supra note 46; Paternoster, Absolute
and Restrictive Deterrence, supra note 2; Paternoster, Decisions to Participate, supra note 2;
Williams & Hawkins, The Meaning of Arrestfor Wife Assault, 27 CRIMINOLOGY 163 (1989).

1990] 669

RONALD L. AKERS

process by which “behavior is acquired or conditioned by the ef-
fects, outcome, or consequences it has on the person’s environ-
ment.” 67 The process involves both reward (pleasant, desirable
consequences) and punishment (aversive, costly, or unpleasant con-
sequences). Reinforcement refers to the process of strengthening
behavior through reward, and the differential reinforcement princi-
ple in social learning refers, in its simplest form, to the process
whereby the acts with the greatest amount, probability, and fre-
quency of reinforcement will be the ones most likely to occur. It
also refers to the overall balance of both reward and punishment for
particular acts or their alternatives:

In a sense the one [behavior] that has been more successful in ob-
taining the desired payoffs will become dominant…. One can be con-
tinued strongly and the other discontinued even more quickly and
effectively if while the first is rewarded the other is punished, even
mildly. 68

Furthermore,
[s]ocial behavior is learned by conditioning, primarily instrumental or
operant, in which behavior is shaped by the stimuli that follow or are
consequences of the behavior and by imitation or modeling of others’
behavior …. Whether deviant or conforming behavior persists de-
pends on the past and present rewards and punishments and the re-
wards and punishments attached to alternative behavior-differential
reinforcement. 69

Thus, social learning incorporates reward and punishment in the
explanation of crime, and the concept of differential reinforcement
applies to the balance of the full range of formal and informal re-
wards and punishment, from the most “rational” calculation of this
balance to the most irrational responses to it. Rational choice does
not add more; indeed, it is more limited.

C. BACKGROUND AND OTHER VARIABLES

To the extent that rational choice theory proposes that only ex-
pected utility affects actions, it disregards the key issue of values or
moral judgments. In contrast, the social learning theory includes
the moral condemnation or acceptance dimension of action under
its “definitions” concept. Even when the perceived material or situ-
ational rewards of taking a particular action outweigh the perceived
costs, if the person has learned “definitions unfavorable” to the act,
he or she is less likely to perform that act. Moral objection to an act

67 R. AKERS, DEVIANT BEHAVIOR (3d ed.), supra note 8, at 42.
68 Id. at 47.
69 Id. at 57.

670 [Vol. 81

SOCIAL LEARNING THEORY

is one type of definition unfavorable to the commission of a crime.
On the other hand, to the extent that one defines an act as desirable,
justified, or excusable, the more likely he or she is to commit the act.
However, he or she is not as likely to commit the act if anticipated
punishment outweighs anticipated rewards. Social learning does
not propose that deviants must hold beliefs that require violation
(although this may occur for certain offenses and ideologically com-
mitted groups), only that there is variation in the extent to which
people believe that they should or should not obey the rule or can
justify its violation. The variation may be in general moral beliefs or
in negative, positive, and justifying definitions of specific acts of
crime and deviance. 70

Some of the empirical rational choice models do include, in ad-
dition to reward/cost, a “moral costs” dimension-conscience, reli-
gious beliefs and commitment, and other moral attitudes and
commitment. These models also lists a range of other variables,
such as social background, upbringing, parental crime, previous
learning, and the influence of friends and other groups. 71 In so do-
ing, these models have broadened the rational choice perspective
beyond the utility function, but they have not gone beyond what is
already proposed by social learning theory. As noted above, the
moral costs dimension is conceptually indistinguishable from the
definitions concept in social learning. Other major variables in the
expanded rational choice models are fairly easily identified as spe-
cific instances of differential association, modeling, definitions, dis-
criminative stimuli, or other social learning concepts.

IV. LACK OF RECOGNITION OF THE RELEVANCE OF SOCIAL

LEARNING THEORY IN RATIONAL CHOICE AND DETERRENCE

LITERATURE

Once stated, most of the foregoing seems self-evident. It also
seems obvious to Pearson and Weiner that deterrence and rational-
ity concepts are partial theories capable of being integrated with
broader perspectives. 72 In their scheme for integrating several the-
ories, Pearson and Weiner subsume formal and informal deterrence
and “utility functions” from economic theory under general differ-
ential reinforcement concepts taken from social learning theory.

70 See id.
71 See THE REASONING CRIMINAL, supra note 2; Paternoster, Absolute and Restrictive De-

terrence, supra note 2; Paternoster, Decisions to Participate, supra note 2; Tuck & Riley, supra
note 46.

72 Pearson & Weiner, Toward an Integration of Criminological Theories, 76 J. CRIM. L. &

CRIMINOLOGY 116 (1985).

1990]

RONALD L. AKERS

Apparently, this has not been so obvious to others. A large body of
literature has developed around deterrence with only some atten-
tion being given to the relevance of general social learning theory.
A rapidly developing body of literature on rational choice, while rec-
ognizing links to deterrence theory and proposing links to social
control theory, has largely ignored the social behavioral perspective
in general, and the social learning approach to criminal and deviant
behavior in particular.

In the introductory essay to The Reasoning Criminal, Cornish and
Clarke make brief reference to the fact that the rational choice per-
spective “recognizes as do economic and behaviorist theories, the
importance of incentives-that is rewards and punishments-and
hence the role of learning in the criminal career.” 73 In the chapter
by Carroll and Weaver, 74 one reference is made to Bandura75 re-
garding cognitive determinants of behavior, but there is no other
reference to social learning theory. Tuck and Riley76 in their section
on “wider criminological theories” make a reference to Akers, 77 and
Elliott,78 and argue that any model of criminal decision making must
include social learning and normative factors. In his review of the
compatibility of rational choice and control theory, Hirschi states
there might be merit in combining compatible theories developed
separately. “A list of such theories would certainly include rational
choice, social control, routine activities, socialization, and at least
some varieties of social learning theory.” 79 These brief and passing
remarks pretty much exhaust the attention paid to social learning in
this key volume on the rational choice perspective. Except for Hir-
schi, himself, the editors and authors show a noticeable lack of
knowledge of Hirschi’s social bonding theory as well.

The inattention by Cornish and Clarke and their contributors to
social learning theory, in spite of its longstanding emphasis on the
rewards/costs dimension in crime (and to some extent their inatten-
tion to social bonding), characterizes other recent literature on ra-
tional choice as well. The same year as the Cornish and Clarke book
appeared, Piliavin, Thorton, Gartner, & Matsueda published an
American Sociological Review article on crime, deterrence, and rational

73 THE REASONING CRIMINAL, supra note 2, at 6.
74 See Carroll & Weaver, supra note 44.
75 A. BANDURA, SOCIAL LEARNING THEORY (1977).
76 See Tuck & Riley, supra note 46.
77 Akers & Krohn, Social Learning, supra note 8.
78 D. ELLIOTr, D. HUIZINGA & S. AGETON, EXPLAINING DELINQUENCY AND DRUG USE

(1985) [hereinafter D. ELLIOTr, EXPLAINING DELINQUENCY].
79 Hirschi, Compatibility, supra note 46, at 117.

672 [Vol. 81

SOCIAL LEARNING THEORY

choice.8 0 They found that the rational choice model tested was only
partially supported, and that rewards related to criminal behavior
more often than did “risk components.” They equate rational
choice with deterrence theory plus a “return component.”8 1

Piliavin et al concluded by calling for a more complex model going
beyond just the reward/cost balance of crime. However, they did
not refer at all to the social learning theory of crime, which already
offers a more complex model incorporating reward/cost balance.8 2

Later, Williams and Hawkins report findings on arrests for wife as-
saults, arguing for an expanded model of deterrence to include both
direct and indirect costs of formal sanctions.83 They link their find-
ings to deterrence, rational choice, and Hirschi’s social control the-
ory in a clear and persuasive way, but make no reference to social
learning. In a footnote in a later paper, the same authors do note
that social control theory is not the only place to find ways of ex-
panding the deterrence concept and refer specifically to social learn-
ing theory.8 4 This is the extent to which the literature on rational
choice and crime has acknowledged social learning theory.

Even in those cases where a “rational choice” model draws di-
rectly on social learning concepts and empirical indicators, the au-
thors are silent about the connection. This is seen most clearly in a
pair of articles by Paternoster, who purports to integrate deterrence
and rational choice into a general social control model applied to
delinquent behavior.8 5 Although his model, like Cornish and
Clark’s model, 6 includes “background characteristics,” he really fo-
cuses on six concepts in his “deterrence/rational choice” model:
(1) affective ties; (2) cost of material deprivation and investments
made in conformity; (3) supportive social groups and opportunities;
(4) informal social costs and informal sanctions; (5) perceptions of
formal legal sanctions; and (6) moral beliefs about specific actions.

What makes this a distinctively rational choice model? Nothing.
Paternoster’s rational choice model is nothing more than a model

80 Piliavin, supra note 2.
81 Id. Klepper and Nagin agreed, and later added “incentives for lawbreaking” into

their deterrence/rational choice model of tax cheating. See Klepper & Nagin, supra note
2.

82 Piliavin, supra note 2.
83 Williams & Hawkins, The Meaning of Arrest for Wife Assault, 27 CRIMINOLOGY 163

(1989).
84 Hawkins & Williams, Acts, Arrests, and the Deterrence Process: The Case of Wife

Assault, (Aug. 1989) (unpublished paper presented to the Society for the Study of Social
Problems, Berkley, CA).

85 Paternoster, Decisions to Participate, supra note 2; Paternoster, Absolute and Restrictive
Deterrence, supra note 2.

86 See THE REASONING CRIMINAL, supra note 2.

1990] 673

RONALD L. AKERS

combining the deterrence doctrine with some social bonding and
social learning concepts. The first concept, affective ties, is simply
attachment taken from social bonding theory.8 7 Costs of invest-
ments in conformity, the second concept in the model, is a direct
application of both the concept of commitment from social bonding
theory and the concept of differential reinforcement from social
learning theory. The rest of the concepts are taken directly from
social learning theory. The third concept, supportive groups and
opportunities, is simply a restatement of the differential association
concept in social learning. The fourth and fifth concepts, informal
costs and sanctions and formal legal sanctions, are both differential
reinforcement. Moral beliefs about specific actions, the last con-
cept, is the same as the concept of definitions favorable and unfavor-
able to delinquent actions in social learning theory.

The conclusion that Paternoster’s deterrence/rational choice
model is really a learning/bonding model becomes even clearer
when we examine the measures of these concepts used in the re-
search to test the model. They are mostly measures used in previ-
ous tests of social bonding and social learning theory.88 For
instance, his measure of informal social sanctions (reactions of
friends and parents) are almost exactly the same as the items Akers
and others have used to measure differential social reinforcement
published in a series of articles over the past decade.8 9 His meas-
ures of moral beliefs also are nearly identical to those that social
learning theorists use to measure definitions favorable and unfavor-
able to criminal behavior. What Paternoster labels “illegal opportu-
nities” is simply the traditional measure of differential peer
association (reported proportion of friends committing the delin-
quent acts) that has been used in delinquency research since the
1950s. Simply taking such measures and re-christening them as ra-
tional choice variables does not make them so; they are still social
learning concepts and measures.

Paternoster did make a brief reference to Akers90 during a dis-
cussion of deterrence in one of the articles91 but he did not recog-
nize any of the social learning concepts and measures from Akers’
article. Furthermore, there is no reference to social learning theory

87 See Hirschi, Compatibility, supra note 47.
88 See, e.g., D. ELLIo’Tr, EXPLAINING DELINQUENCY, supra note 78; Akers & Cochran,

Adolescent Marijuana Use, supra note 22; Krohn, Community Context, supra note 23.
89 Akers & Krohn, Social Learning and Deviant Behavior, supra note 7; Akers & Cochran,

supra note 22; Krohn, Community Context, supra note 23; Krohn, Social Learning Theory and
Smoking, supra note 23.

90 Akers & Krohn, Social Learning and Deviant Behavior, supra note 8.
91 Paternoster, Absolute and Restrictive Deterrence, supra note 2.

[Vol. 81674

SOCIAL LEARNING THEORY

in either Paternoster article. Hirschi’s social bonding theory92 fares
only slightly better in the two Pasternoster articles. Paternoster
found that the threat of formal legal sanctions is less important than
the informal social sanctions and “nonlegal consequences.” The
strongest independent variables in his research are differential peer
association and moral beliefs. These are exactly the kind of findings
predicted by social learning. Paternoster had earlier reported very
similar findings for a very similar model.93 At that time the authors
made no reference to either the rational choice or economic models,
but they did discuss the relevance of social learning and social bond-
ing theories to deterrence. The data, measures, and findings are
basically the same in the earlier and later articles. What happened
in six years to transform the same variables from a deterrence model
explicitly related to social learning and social bonding theory into a
rational choice model for which bonding has little, and learning has
no relevance?

V. CONCLUSIONS

Assumptions about the level of rationality in criminal acts do
not distinguish rational choice from current criminological theories.
The basic ideas and central propositions of deterrence and rational
choice theory as currently applied in criminology have already been
captured in the social learning approach to deviant and criminal be-
havior. Specific measures and application of these principles in re-
search on rational choice models may be different; the concepts are
not. In some of the most recent rational choice literature, such as
Paternoster’s, even the empirical measures do not differ. Social
learning theory incorporates concepts and processes which the nar-
row rational choice models do not. When broader models of ra-
tional choice have been developed, they begin to take on the
appearance of social learning models.

By the time that rational choice models began to take hold in
criminology, there already had developed a rich body of theory and
research on crime and deviance within the social behaviorist tradi-
tion, a tradition which already had incorporated the central proposi-
tion of rational choice theory. Yet, none of that tradition was
consulted by proponents of rational choice theory. Rather, eco-
nomic theory was imported and modified as rational choice models
of crime. These models then were referred to in modifying the de-

92 Hirschi, Compatibility, supra note 47.
93 Paternoster, Perceived Risk and Social Control, supra note 19.

1990] 675

RONALD L. AKERS

terrence doctrine in criminology, as if none of the behavioral tradi-
tion existed.

One may wish to propose that what seem to be obvious theoret-
ical links disappear upon closer examination, and that rational
choice offers a brand new approach with concepts and propositions
that differ significantly from anything in social learning or other ex-
tant criminological theory. Such an argument would be difficult to
sustain however, and no one has yet attempted it. Instead, the issue
is simply ignored. The links of deterrence and rational choice ex-
planations to social learning principles are clear and cannot be ex-
plained away. Social learning is an established, well developed, and
well researched theory widely known in criminology. Therefore, it
is incumbent upon rational choice theorists to show how their
“new” models do or not relate to it. They should also examine
carefully other existing criminological theories instead of relying on
blanket characterizations of “traditional criminology.” Reinvention
of the wheel should be avoided even in criminological theory.

676 [Vol. 81

Journal of Criminal Law and Criminology

Fall 1990

Rational Choice, Deterrence, and Social Learning Theory in Criminology: The Path Not Taken

Ronald L. Akers

Recommended Citation

Rational Choice, Deterrence, and Social Learning Theory in Criminology: The Path Not Taken