The Oxford Research Encyclopedia of Criminology & Criminal Justice will be available via subscription on April 26. Visit About to learn more, meet the editorial board, or recommend to your librarian.

Show Summary Details

Page of

 PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, CRIMINOLOGY AND CRIMINAL JUSTICE ( (c) Oxford University Press USA, 2016. All Rights Reserved. Personal use only; commercial use is strictly prohibited (for details see Privacy Policy and Legal Notice).

date: 20 April 2018

Control Balance Theory

Summary and Keywords

Control balance theory (CBT) was developed in the mid-1990s, primarily to illustrate a particular method for building integrated theory and to show how general theories are useful in addressing various issues in studies of crime and deviance. A major theme on which the theory has been built is the idea that deficient control, a well-established classic notion, and excessive control all can have deviogenic consequences. In addition, the theory rests on expectations that sufficient explanation will necessarily involve complex arguments. Thus, CBT not only attempts to explain the phenomena within its domain, but it also challenges simplistic theories, contentions that theoretical integration is both impossible and undesirable, and neglect of contingencies in theorizing.

Keywords: theory, Provocation, control ratio (CR), explanation, deviance

Control balance theory (hereafter CBT) was introduced in 1995 and refined in 2004 to take account of research and criticism. It was meant as an example, using a particular technique of theory building, of how to create an integrated explanatory system applying universally to all instances within a claimed domain. CBT attempts to assimilate a diverse array of focused theories into a larger, more comprehensive account capable of answering questions of “why” and “how” concerning any deviant behavior. Responses to criticism and evidence highlight the interplay between intellectual arguments, research and criticism, and feedback, ideally leading to theoretical modification and improvement needed for advancement.

The domain of CBT includes conduct generally regarded as inappropriate to members of a particular group or a collectivity of groups. Deviant acts are those thought of in a pejorative light (wrong, bad, inappropriate, etc.) by members of a given group, or that “typically” attract a negative reaction from members of the group or elicit sanctions from authorities acting on behalf of a collectivity. A typical negative reaction may involve spontaneous social actions, such as shaming, labeling as bad or wrong, or social rejection, as well as punishing actions (such as incarceration, which is imposed by those who have the authority granted by groups such as governed societies, churches, voluntary associations, or businesses to impose sanctions).

Since CBT treats deviance as relative to a group or cultural context, it encompasses all forms of misbehavior, from trivial to very consequential, that can emerge from the norms of a group or the general norms of a collectivity of social entities such as business organizations or nation-states. CBT’s explanatory domain includes criminal acts and the rates of such acts, although some criminal acts may not be explicable because, within the context being considered, those crimes are not deviant and therefore not liable to be countercontrolled should someone or some entity attempt to use them to gain some control. Importantly, however, the explanatory domain of CBT is not limited to crime but includes a wide range of noncriminal acts, such as someone betraying a friend, an employee using company equipment for personal purposes without permission, or a teacher denigrating a student in public.

To illustrate the general sweep of the theory, it is instructive to consider some actual uses to which CBT has been put. The following examples identify instances where research scholars have tried to explain manifestations of deviance within the framework of an empirical test of hypotheses from CBT, or have used the concepts and causal arguments of CBT to interpret or understand incidents of misconduct. The applications include street crimes (Baron & Forde, 2007; Delisi & Hochstetler, 2002; Fox, Nobles, & Lane, 2016; Higgins, Lauterbach, & Tewksberry, 2005; Hughes, Antonaccio, & Botchkovar, 2015); intimate partner violence (Castro, Nobles, & Zavala, 2017); college exam cheating (Curry, 2005); drunkenness and marijuana use (Curry & Piquero, 2003); workplace crime (Dunaway et al., 1999); criminal sex offending (Wood & Dunaway, 1997); unconventional sex practices (Piquero & Hickman, 1999); unhealthy dieting (Hickman & Piquero, 2001); blowing the whistle on fellow police officers violating rules (Hickman, Piquero, Lawton, & Greene, 2001); exploiting another student’s class notes (Higgins & Lauterbach, 2004); stalking (Nobles & Fox, 2013); exploitative corporate crime (Piquero & Piquero, 2006); rogue securities trading (Rafeld, Fritz-Mortenthal, & Posch, 2017); computer misuse (Williams, 2008); and even criminal victimization (Nobles & Fox, 2013; Piquero, 2003). Moreover, it has been employed to derive potential social policies for controlling criminal misconduct (Tittle, 2010).

As broad as the applications have been, they still do not show the full potential applicability of CBT. After all, it is a sociological truism that sustained human interaction spawns norms to express and regulate that interaction. Where norms emerge, violations become a possibility, and nonnormative behavior invites countercontrol. Since CBT claims the ability to explain misconduct wherever and whenever it occurs, there remains a rich thicket of deviant acts yet to be considered within this model.

In addition, all the research focus on CBT so far concerns explanation of behavior by individual persons, often referred to as microlevel explanations. Yet, a competing and often favored explanatory focus of scholars, especially sociologists, is macrolevel, sometimes referred to as “ecological” (see Baumer & Arnio, 2016). Ecologically oriented scholars are interested in aggregations of individual deviant acts, such as homicide rates, rates of domestic abuse, or rates of tax evasion, and sometimes in more fundamental social and organizational expressions of cultural norms, such as structural racism, notions of honor, or extreme so-called masculinity. Although CBT lends itself to the macrolevel as well as the microlevel, so far (to my knowledge), there have been no attempts to examine, for instance, if degrees of average control imbalance among ecological units such as states or nations predict different rates of deviance. Further, there have been no efforts to explore how overall levels of racism or other cultural expressions might correspond with particular levels of control imbalances. Finally, CBT’s potential for explaining conflicts or group exploitations such as unethical business practices, international conflicts, or group exploitations of one kind or another has been overlooked.

Research emphasis on the microlevel is certainly consistent with CBT’s focus on decision-making and behavior by individual persons. Still, because of perennial social science difficulty in handling differences between explanations of individual conduct and explanations of rates of misconduct or actions of organizations, or in developing truly cross-level theories (Messner, Krohn, & Liska, 1985; Short, 1998), one of CBT’s main virtues would seem to be its presumed ability to account for both microlevel and macrolevel phenomena.

Thus, CBT addresses and offers a pathway for solving several key issues in social science. In addition to illuminating a method for integrating disparate theories or theory fragments and in permitting explanations of microlevel and macrolevel phenomena within the same theoretical model, CBT offers a common account of misbehavior by those with high control ratios (CRs, sometimes called control surpluses), as well as those with low CRs (previously described as control deficits), often thought to be akin to socioeconomic status (SES) differences. Traditionally, a major cleavage among social scientists turned on a supposed negative association between SES and crime (Braithwaite, 1981; Tittle, Villemez, & Smith, 1978, 1982), with almost all sociological theories offering explanations for this imagined truism (Tittle, 1983).

While that putative association is probably no longer regarded as a truism (Agnew, 2016), many social scientists still believe that SES is a key predictor of crime, presumably because they assume unlawful behavior to be motivated mainly by material deprivation and need. CBT, however, suggests that the causes of crime and other misbehavior have to do with control—trying to get more of it and trying to avoid having too little of it—and emotion-generating situational responses concerning it, features that are only distantly and indirectly linked with material means and needs. In short, CBT offers a remedy for the SES divide. A student of social behavior no longer needs to assume that the social world spins on the axis of SES.

Finally, CBT accommodates explanatory notions rooted in past experiences, current influences and affiliations, and contemporary situational characteristics, treating all three as essential to understanding deviant behavior. For example, general CRs, the key variable implied by CBT, are somewhat stable throughout life at the same time that they are subject to developmental change. However, situational CRs are highly variable from individual to individual and for any given individual from situation to situation. Therefore, careful application of CBT will reveal that one’s “background” (i.e., things influencing global CRs) is part of the equation explaining deviant behavior, just as general features of present global CRs and control characteristics of the immediate situation are (Wikström & Treiber, 2016). The theory, then, is intended to recognize long-standing issues in studies of deviance of various kinds and to build into itself solutions to those issues.

How well CBT accomplishes its multiple goals may be debatable, but its attempts and hints of eventual success sustain its followers.

The Theory

The Basic Model

CBT purports to explain the probability of some kind of deviance occurring, as well as the type of deviance, expressed in a restricted range on a scale of control balance desirability (CBD), which is ultimately committed—two related but distinct objectives. Deviant acts are themselves characterized by their simultaneous ability to produce long-range and effective gains in control and the extent to which an entity can engineer the act without personal public involvement. These two features of potential deviant acts converge to define CBD.

The CBT model contends that misbehavior results from a sequence of “decisions,” by those with control imbalances who are provoked into an acute awareness of their control circumstances. That chain of decisions eventuates in the commission of one or more deviant acts from among those that are possible and likely in the situation; that is, within a similar zone of CBD predictable from the CR and other variables in the CBT scheme.

The theory assumes that people (this also implies other social entities) desire autonomy and are motivated to seek control of their social and physical environments. The key variable of CBT, then, is the CR. All individuals can be characterized by the amounts of control that they exercise relative to that to which they are subject, both generally (global CR) and situationally. Those who are relatively balanced (somewhere near zero) are theorized to conform most of the time, for reasons that will become clear as the description of CBT proceeds. Those with exceptionally low CRs (i.e., the control to which they are subject exceeds to the extreme the control that they can exercise) are liable to be submissive. Such oppressive control renders its victims unable to imagine anything but obedience. All others are characterized as having unbalanced CRs.

The original version of CBT (Tittle, 1995) differentiated positive from negative imbalances, with particular kinds of deviance theorized as being produced by control surpluses and other forms being generated by control deficits. In the refined and presently operative version of CBT (Tittle, 2004b), a differentiation is no longer made between surplus and deficit imbalances. Instead, one’s CR is either low or high—actually, some gradation on a continuum. That position influences the likelihood of provocation, the nature of provocation, and, if the variables of the model play out, the expression of that likelihood in some form of deviance generally, or in the range of the CBD continuum from whence a specific deviant act is selected.

The deviance-generating process begins when an individual with a control imbalance (either negative or positive) is strongly reminded of that imbalance by some stimulus, called a provocation. Many things can remind a person of an imbalance. Among those with low CRs, provocations might include receiving a bill or threatening letter, being denied entrance to a show or event, being fired or scolded by a boss, being insulted or put down, being cut off in highway traffic, and many others. Provocations for those with high control surpluses take the form of denial, blockage, or challenge to the superior control that the person thinks she has. For example, a teacher may have a high situational CR because of her positional influence over numerous students and other teachers who regard her as a role model, but she is subject to provocation by perceived disrespect from students, scolding from parents, or harsh criticism from her principal. Similarly, a street gang leader has a high situational CR by virtue of his influence over other gang members and respect that he garners from rival gang members or the police. Yet, this elevated situational CR creates a liability for provocation by challenges from other members of his own gang or by rival gangs. Or, a wealthy entrepreneurial woman with many employees and much political influence may have a generally high CR, but because of that, she is likely to be provoked if she has to stand in line at the grocery, overhears one of her employees making derogatory remarks about her to other employees, or is fired from her company’s board of directors.

Provocations are unpleasant—indeed, sometimes devastating and humiliating. Provocations not only remind one of his or her control situation, but they usually carry an emotional sting; in fact, the result is often feelings of having been slighted, and sometimes even rage. They inspire their recipients to want to do something about the control imbalance, the provocation, and/or the humiliation. Therefore, it is the case that control imbalances set the stage for provocations, while provocations stimulate a search for some way to gain back the control perceived to have been lost or to extend the control that one has in order to overcome the vulnerability to feelings of demeanment. Having been stimulated to do something to improve her control situation, the person is likely to contemplate deviant behavior of some kind. Contemplation of deviant acts is highly likely because deviance is usually the easiest and most effective way to extend one’s control.

Given a desire to fix an unpleasant situation, and having realized that deviance may hold the answer, the individual begins to consider whether various norm violations can be accomplished, especially whether particular forms of misconduct are possible for him or her. An act that is feasible for a particular individual represents an “opportunity,” but deviant acts that are possible are not necessarily effective for the purpose at hand, and they may involve much potential countercontrol, especially if the individual must be personally “hands-on” instrumental in their accomplishment. For example, a customer might wish that she could buy out a business in order to fire a clerk who she perceived insulted her, but if she has no resources to make such a purchase, then there is no opportunity to fire the clerk. She also might contemplate disabling the car of the clerk or one of her supervisors, but without mechanical knowledge of how to do that, she has no opportunity. Or, she might consider assaulting the clerk or somebody else. Assault is certainly possible or opportune, but its effect for gaining control is limited (she may be incarcerated for it; she may be barred from retail establishments in the area in the future), and the clerk or other victim may retaliate by hitting her back. Opportunity to commit a deviant act to gain control or to overcome humiliation, then, only means that the action is possible; it does not mean that the act is capable of accomplishing its intended purposes.

Contemplating various potential deviant acts as vehicles for gaining control involves balancing the gain in control against likely countercontrol that might be activated by commission of each of these acts. An individual may gain a lot of control by stealing a store’s payroll, but the state may respond by taking his freedom, so overall, he is better off tolerating the control being exercised against him or, more likely, finding some other type of deviance that has less chance of attracting such disadvantageous countercontrol. Ultimately, deviant conduct emerges when the individual or social entity decides that the gain in control from one or more possible deviant acts outweighs the potential loss in control associated with those acts.

In the control-balancing process, the person may contemplate any number of deviant acts. Some of them carry greater chances of enhancing control than others, and some of them imply greater countercontrol than others. Since deviant acts are arrayed on a continuum of CBD, one can say that the person slides up and down that scale within a range constricted by that individual’s CR and certain other factors, searching for the most appropriate deviant expression. Thus, if assault is not workable, then one might choose some form of vandalism, and if vandalism is not a good choice in the situation, a person might resort to profane insults. Also, if the direct source of one’s bad feelings or control position potentially spawns too much countercontrol, the deviance-seeking individual may try to extend his or her control in other contexts.

This brief statement outlines how and why deviant behavior occurs—that is, to extend control and to overcome unpleasant emotions such as humiliation in the face of events that remind one of one’s low CR (or calling attention to the failure of a high CR to be recognized). However, clarification of concepts and arguments, as well as cautionary insights, is essential. The following section elaborates the model, the concepts, and issues.


Control Ratios

CRs represent the degree of control over various aspects of the world (other people, physical things, and circumstances) that an individual has relative to the control that various aspects of the world have on her or him. The concept is easy to imagine but hard to operationalize. For one thing, sources of control, as well as vulnerabilities to control, are extremely varied, ranging all the way from friendships to wealth and including both personal and social characteristics. There are literally dozens (if not hundreds) of things that can come into play to enhance or detract from one individual’s or one entity’s CR. In addition, a CR depends on the domain at a given time. There are huge variations sometimes between a person’s control situation in a work environment and her home or in an alleyway, with an armed robber threatening her. Furthermore, CRs, whether applying to broad swaths of life or to focused situations, change rapidly. With a work demotion, a person can almost immediately shift several points on a scale of relative control. In a short time, a wealthy woman can move from a high to a low CR as she goes home to an abusive husband from a commercial establishment where her money may attract much attention.

Technically, then, any individual or social entity has numerous CRs. In order to realize the full power of CBT, one would have to examine CRs for individual entities at multiple time points and for numerous domains. However, such precision is neither likely nor necessary for useful application of the theory. Although there is substantial variation in CRs over time and by domain, there is also likely to be much typicality that expresses the average CR that people possess. Some researchers, for instance, have employed measures of global CR, sometimes made up of many domain-specific CRs, and others have derived general or global CRs without differentiation of domain-specific CRs.


The higher the CR, the more likely a person is, when provoked and motivated to commit deviant behavior, to choose deviance high in CBD. However, that generalization, like all others in CBT, is subject to various contingencies. For example, even though the relationship between the CR and the CBD is basic, it can be influenced or diverted if the person has low self-control (Gottfredson & Hirschi, 1990). One with high self-control can overcome the desire to strike out impulsively in response to a provocation, allowing the CBT process to unfold fully, while those with low self-control may quickly act to regain their perceived lost control without effectively control balancing, or perhaps even considering the possibility of countercontrol.

Similarly, some individuals may contemplate a deviant act in response to a provocation but be unable to carry it out because of personal moral constraints. Indeed, morality may prevent one from even imagining the possibility of certain deviant acts (Brauer & Tittle, 2017; Wikström, 2006). As a final example, one may become provoked and contemplate some form of deviance but not act because her disapproving friends and associates constrain her.

Overall, then, although CBT spells out a causal process for deviance, the fullness with which that causal process unpacks is subject to many conditions. Therefore, in applying the theory or testing it, one must be careful to bring into consideration circumstances under which its variables are theorized to operate with greater or less strength.


While the deviance-generating process has been described here thus far as if it were completely rational, with careful weighing and deciding step by step (Piquero, 2002), in reality, the process is often quite ragged. Neither people nor other social entities deliberate much. Usually habit, impulse, or routinization is at play, and frequently the full control-balancing sequence occurs almost instantaneously, without much evidence of rationality at all. Further, humans and other social entities tend to accommodate themselves to control circumstances, bringing their control imbalances to the fore of consciousness only when provoked.


Provocations are events, actions, or circumstances that acutely remind those with unbalanced controls of that fact and its possible consequences. Insults, for instance, remind people with low CRs that they cannot control things affecting their lives, while they remind people with high CRs that the deference to which they think they are entitled is not as automatic as they assumed. In both cases, those persons are likely to experience some negative emotions (Agnew, 1992, 2007)—often humiliation, or feelings of having being “put down”—and as a result, they are likely to ponder ways to gain in control to compensate for the insult or to overcome the feelings of humiliation. Being fired from a job is another reminder to those with weak control balances that they are subordinates, while losing a contract may remind those with strong control balances that they are not so powerful after all. Here, too, the urge to gain control is likely to stimulate thoughts of deviant behavior. Deviant motivation is simply the realization by an individual that deviant behavior can help rectify the provoking or humiliating situation or the control imbalance that is recalled by provocation.


When people are provoked and come to realize that deviant behavior can be helpful in gaining or restoring control, they cannot necessarily contemplate committing just any deviant act. Some are simply not possible, either because they cannot be accomplished physically in a given context or because the person’s circumstances (implicated in the CR) will not permit them. A boss can contemplate firing a worker, but that worker cannot contemplate retaliating in kind. Similarly, a worker may not be able to contemplate using a piece of company equipment (such as a computer) for his own purposes if he cannot operate it, while a boss may not be able to imagine beating up a physically superior worker. The physical possibility of various deviant acts (opportunity) is a crucial, essential element in the chances of deviant behavior occurring.

Control-Balance Desirability

In addition, not all deviant acts are realistic for a provoked individual to contemplate. Deviant acts usually involve harm or inconvenience to others or they tweak moral concerns. That is, any deviant act carries some potential for countercontrol, and sometimes countercontrol is so likely, and so potentially constraining, that any gain in control from a given deviant act would be threatened by the consequences of the act. Provoked/motivated entities must engage in control balancing, in which they weigh the gains in control from deviance against the potential loss in control that might be brought about by the deviant act.

However, a provoked person is likely to think first of opportune deviant acts that have immediate consequences and that involve direct, hands-on action by the person, such as punching, vandalizing, or obstructing. Such acts are likely to be the most satisfying emotionally, but they also are most likely to bring about strong countercontrol, so they are less effective in the long run for gaining control. CBD is a characteristic associated with a deviant act; it incorporates the chances and degree to which that act will have long-range effects in altering control imbalance and the degree to which the act requires personal direct involvement by the perpetrator.

The most control balance desirable acts have the maximum long-range impact on altering control imbalances, and they are the most impersonal (i.e., they can be accomplished without direct, hands-on action). For example, a powerful corporate executive might be able to damage an insulting rival’s economic future without being obviously and directly involved, thereby gaining a lot of control but avoiding the costs that would have accrued had he or she directly yelled at or punched the competing executive. The higher the CR, the greater the potential for engaging in more control balance desirable acts. And, as noted before, the degree to which control balancing is conducted rationally depends on the self-control of the individual.

Hydraulic Relationships

CRs are complex phenomena, made up of relative controls of all kinds in numerous domains, with both global and domain-specific control manifestations. For instance, most people are sometimes controlled by weather conditions, traffic delays, and the actions of foreign and domestic governments. Further, control is, at least partly, dependent and reciprocal. Part of any person’s control stems from others having less control; if one individual gains or loses control, it will be to some extent because somebody else or some other entity correspondingly loses or gains it. Therefore, CBT implies many reciprocal or feedback relationships. A particular CR may lead to deviance, but that same deviance may change the CR that caused it.

Balanced Control

According to the CBT theory, people with balanced CRs are least liable to offend because they are less vulnerable to provocation, less likely to become motivated toward deviance, and more likely to face countercontrol for deviant actions that might be undertaken to change their control situations. This does not mean that such people will never offend, only that their probabilities of deviance are lower than for others. Those with balanced CRs do encounter provocations, but those provocations, while emotionally painful, have small chance of leading to deviance.

The Autonomy-Desiring Assumption

The cognitive, or decision-making, sequence of CBT assumes that people desire autonomy; indeed, as originally presented, the theory contends that the desire to control one’s social and physical environment is more or less universal, present already in infants who seek attention and care by crying, and later delight in influencing adults by knocking things off tables. This assumption was put forward as a temporary one of convenience—a theoretical tool, as it were—to help simplify a complicated set of causal interconnections. It is quite likely that the desire for autonomy, while present in all or most people, actually varies in strength (Burger, 1992). It is my hope that someone will later reformulate CBT with that possibility built in, spelling out causal interconnections under various degrees of actual desire for autonomy. Until that reformulation has been accomplished, CBT will probably not achieve its full potential.

Is CBT a Revenge Theory?

It is important to understand that when an individual with a control imbalance is provoked and goes through the decision-making sequence to arrive at one or more deviant acts that he or she might commit, those deviant acts are not necessarily ones focusing on the perceived source of provocation, nor is the one eventually selected. Some of the examples given so far do involve retaliatory acts against those who have called attention to a control imbalance. However, those instances are not necessarily definitive. A provoked person seeks some way to gain more control, and those means are as likely to implicate innocent or irrelevant parties as they are to be aimed at the true source of the agony. Those who are rudely fired, for example, may be as likely to abuse their children as to seek revenge against a supervisor or a company. Those expressing road rage may be as likely to force an innocent motorist off the road as they are to try to sideline a large truck that they perceive to have done them wrong.

Not only are the objects of deviance often not the same as those who provoked the deviance, but many provocations are inspired by nonhuman entities. People are sometimes reminded of a control imbalance by the weather, by traffic, by movements of interest rates, or by accidents. It is, of course, absurd to imagine that a perception of powerlessness could be relieved by acts of deviance directly against such phenomena. So CBT is not a revenge theory, although it may help us understand those cases that actually reflect revenge.


Empirical Standing

For maximum usefulness and credibility, any social science theory must meet at least two sets of expectations. One has to do with accuracy or validity; that is, does the theory actually permit an explanation of the things that it claims to explain? Since scholars disagree about the meaning of explanation, they usually employ an empirical criterion of predictive power to answer this question. The gold standard for judging a theory is its reputation among social scientists, as indicated by the number, sophistication, and variety of successful tests credited to it using data from the world of real people and events. Testing a theory typically involves the logical derivation of hypotheses concerning how two or more concepts from the theory are linked, followed by operationalization of the concepts into empirical variables that can be observed and manipulated to determine if the predictions implied by the hypotheses hold up.

Thus, if a theory implies that A leads to B, then scholars should observe that, within the data examined, A and B are in fact linked in various ways consistent with the hypothesis. If this turns out to be the case, the theory gains credibility—provided that critics perceive that the hypothesis was accurately derived, the variables were reasonably measured, and the results were correctly interpreted. On the other hand, if tests fail to confirm the hypothetical predictions, scholars may lose confidence in the theory in question. Since theories usually yield many potential hypotheses, some more sophisticated and complex than others, and each hypothesis can potentially be addressed with various kinds of data using multiple methods of measurement and analysis, no particular test is regarded as the final word. Theories are judged as more or less credible, depending on the scorecard from the tests that have been performed.

However, there are many potential problems with empirical testing. Inaccurate derivation of hypotheses is common; data are often flawed; measurement is frequently questionable; and results are sometimes complicated and subject to alternative interpretations. Furthermore, from time to time, popular sentiment in favor of (or against) a particular theory sustains a reputation for it that is not actually justified by the collective body of empirical evidence. Overall, there seems to be considerable generosity in expressed judgments of empirical validity, with practically all theories being supported under at least some conditions. Indeed, sociological criminology might be characterized as a “zombie” science, filled with theories that, based on the available empirical evidence, are dead even though they are still included in the lexicon and meticulously set forth in textbooks and anthologies.

In addition, certain conventions in the social scientific community, though demanding at the testing level, work against a rigorous overall evaluation of a theory. The first of these is reliance on tests of significance to determine if a hypothesis or a part of a hypothesis is empirically supported. Tests of significance are popular because they provide a statistical rationale for judging whether a given numerical outcome is likely to have been a chance occurrence, thereby helping to rescue research from subjectivity. Unfortunately, almost universal employment of significance tests also means that in practice, the criterion of theoretical power turns out to be simply a better-than-chance prediction, not the magnitude of effects observed, the interconnection of results, or the sophistication of causal processes laid out by theories.

A second constraining feature of social science practice is an affinity for simple explanations. Although almost all observers of any social scene easily notice complex causal networks, feedback relationships, the operation of multiple contingencies, and numerous other deviations from straight “A leads to B,” simple theories inspire the most research attention and usually win the contest for empirical credibility when pitted against theories rooted in more complex interlinked causal mechanisms.

And why shouldn’t that be? Testing complex theories is difficult, time consuming, and fraught with uncertainty about how the unique approaches required will be judged by journal reviewers. Providing acceptable, much less convincing, evidence necessitates the formulation of complicated hypotheses because of multiple causal paths and potential contingencies, the use of innovative variable measurement because of new concepts not readily indexed from data archives, and sometimes consideration of unconventional criteria of empirical accuracy. Where careers depend on research publication (which often is measured by sheer volume), it is no surprise that many will choose to focus on simple theories rather than complex ones. This is not to say that such a choice is wrong, only that judging theories by collective bodies of empirical results is not as simple as it might seem.

The Evidence

CBT has not been a popular theory, at least as judged by citation and attention from potential testers. In the 22 years since its conception (Tittle, 1995), not even one test or application per year has been published. However, as was mentioned in the introduction to this article, the extant tests range over a broad spectrum of dependent variables. Moreover, they employ a variety of methods, both for tapping deviance and for indexing other variables in the theory, such as CRs. The most popular means for measuring the dependent variable apparently has been the vignette technique, although survey and observation have been used as well. For obvious reasons, the self-report questionnaire has been the main source of information about perceived controls, both exercised and experienced.

Perhaps the most thorough review of this limited research literature concerning CBT is that by Fox, Nobles, and Lane (2016). They observe that the supportiveness of research on CBT depends somewhat on the source of the data used. Although no test has reported results fully consistent with the derived hypotheses, almost all of the empirical studies have found some support under some conditions. Hence, the overall empirical status of CBT can probably best be characterized as “promising but guarded.” However, the conclusion from what seems to be the most sophisticated study so far is substantially more positive (Fox et al., 2016):

The evidence illustrated in this study suggests support for Tittle’s (2004b) basic assertion of a moderating relationship between control imbalance, low self-control, and deviance (via control balance desirability)” . . . “the results from this study add a new dimension to boost confidence in control balance as a general theory of crime” . . . “In the end, this study illustrates the utility and generality of control balance theory for explaining offending among a large sample of recently incarcerated men and women.” (p. 948)

Thus, the evidence compiled so far is not as strong and compelling as CBT proponents might desire, and it reveals some persistent problems. Still, research results seem to confirm that CBT is on to something, even if the details are not yet fully ironed out. This conclusion appears especially justified because the extant research concerning CBT probably underestimates its worth. There are at least three reasons for believing that conclusions about CBT drawn from the existing body of research are too conservative.

First, almost all the extant CBT tests involve predictions of specific dependent outcomes, one at a time. But CBT argues that those who experience the causal impetuses laid out by the theory actually try out in their minds a number of different forms of deviance, ultimately selecting one among a set of potentials. The set from which an individual chooses includes deviant acts that are similar in the extent to which they provide long-range gains in control and permit the avoidance of direct personal involvement in the act of deviance. However, deviant acts falling into a narrow zone of this overall continuum of CBD, from which a motivated person may select a deviant act to gain some perceived control, do not necessarily appear similar in other ways besides CBD. Thus, the theory does not anticipate that process A necessarily leads to act B; rather, it implies that process A leads to either B, C, D, E, or F, all of which are similar in their CBD weights even though they are in fact different types of behavioral acts.

Hence, research serially employing single, specific, dependent variables inadvertently underestimates the potential power of CBT. For example, suppose that in a given study, under CBT-specified conditions, a given subject projects, admits, or in some other way shows an inclination toward deviant act D. If the researcher’s preselected dependent variable in this hypothetical study is B, then the data will register this individual as an incorrect prediction because she did not commit act B. But, in fact, the case would have been an unrecognized correct prediction because the theory contends that the dependent variable is (a) some form or forms of deviance from among many possibilities, or more specifically, (b) one or more forms of deviance with a similar CBD score (that is, within a predicted range of CBD, including B but also including C, D, E, and F).

To provide more faithful tests, researchers must do one of two things. First, studies should allow subjects a large and varied choice of deviant acts as possible responses to the conditions theorized by CBT. Second, and more appropriately, researchers could attempt to establish the CBD scores for a fairly large number of deviant acts so that those acts could be arrayed on a continuum that permits similar CBD acts to be identified. Having established such a continuum, a researcher’s task would be to test the extent to which CBT models anticipate one or more acts that are similar in degree of CBD, as well as the extent to which the predictions behave according to expectations under various contingencies and conditions specified by the theory. The greater the number of CBT conditions that are accommodated in a study, the narrower the range of CBD within which predicted deviant outcomes can be made. In other words, the more inclusive and faithful the test, the closer it should get to predicting a single dependent variable, as reflected by a very restricted zone of CBD.

Two recent studies do make an attempt to accommodate the idea that those propelled by CBT conditions are liable for some form of deviance, even though they may select any one (or more) from a set arrayed over a restricted range of CBD. Fox et al. (2016) and Hughes, Antonaccio, and Botchkovar(2015) use as their dependent variables “variety” indexes composed, in the former, of 13 crimes, and in the latter, of 6 types of misconduct. In these two studies, a CBT actor has several options for expressing herself rather than the one or perhaps two permitted by other studies. Although an important improvement, these particular variety indexes do not completely fulfill CBT requirements. Even having up to 13 violations does not approach the full range of deviant choices available to individuals within the sample studied (in particular, the list omits all noncriminal deviant acts).

Not only do Fox et al. (2016) and Hughes et al. (2015), and other studies before them, fall short in their choices of dependent variables, but none of the research so far has actually taken the CBD of the deviant acts treated as dependent variables into account. Of course, some of this research was done before the theoretical refinement of 2004 appeared, setting forth the notion of CBD. Some of the contemporary research reports do mention CBD and acknowledge its importance, with Hughes et al. (2015) even attempting a single-item measurement of CBD. Unfortunately, that item does not get at one of the two major elements of CBD—impersonality of the act. Furthermore, Hughes et al. (2015) were not able to bring variations in their indicator of CBD into their analyses in a meaningful way.

It is true enough that a weak test of CBT can be performed without considering CBD. Control imbalances should predict some form of deviance, so long as research subjects have a wide choice of acts, and the more the conditions of CBT are fulfilled (such as provocation or frequency of provocation), the greater the chances are of predicting some form of deviance. However, stronger hypotheses involving the prediction of specific deviant acts are not justified without taking into account the CBD over an array of deviant acts. If future research brings CBD into the data gathering and analysis, researchers should anticipate not only full confirmation of the weak hypothesis, but also confirmation of stronger hypotheses that narrow the range of the CBD of the dependent variables.

A third reason to imagine that the empirical strength of CBT is underestimated by extant research is the absence of efforts to estimate reciprocal, or feedback, effects. The theory portrays deviant behavior as a maneuver by a social entity to overcome a problem. In this sense, CBT is a type of strain theory (Agnew, 1992). Deviance is intended to (and presumably sometimes does) enhance control for the social entity undertaking it at least temporarily, but often permanently.

Imagine, for example, that individual A, with a general CR of −15, encounters a provocation on her job (perhaps her boss yells at her), which produces feelings of having been diminished, put down, or humiliated, reminding her that she has little control of conditions of her life. She wants to do something about it and realizes that she has several options for trying to even the score—some more realistic than others in terms of opportunity and potential countercontrols. In the control-balancing process, she weighs her options, considering long-range gains in control that she might realize, her opportunity to commit various acts of deviance, her ability to accomplish various forms of misconduct, and potential countercontrols. Of the available acts that she might commit, she decides on vandalism, lingering after work to stuff waste paper into one of the company’s toilets, perhaps even the boss’s private toilet. This act makes her feel somewhat more powerful, and when other workers learn of the defiance, she gains a measure of prestige among them. That prestige, in turn, yields additional control in the form of free lunches, friendships, and help with work problems.

Now imagine that this individual is participating in a survey by a social scientist interested in assessing CBT and is answering questions in order to permit the scholar to establish her CR and its interconnection with deviant conduct. The interview (or response to a questionnaire) is at T1 shortly after the provocation, so the data show a CR of −20 instead of the usual −15 experienced by this person. This degree of control imbalance indicates a strong likelihood of committing one or more deviant acts. However, if the data collection is at T2, after she has committed the vandalism and experienced the rise in her CR, her overall control deficit may be measured as −10, suggesting that at T3, her chances of some deviance are somewhat lower than they were at T1. Thus, in a one-shot survey, outcomes may depend on where in the causal chain the data are collected, being at the causal end for some subjects and at the outcome end for others, and with no means of assessing reciprocity. And even if data were collected longitudinally, they would not necessarily allow reciprocal effects to be assessed because collection points would still cut into numerous reciprocating causal chains in inaccurate ways.

In summary, it appears that CBT enjoys some support in the research literature, although the level of confirmation may not be sufficient to persuade critics or satisfy avid fans who might have expected far more confirming outcomes. Those who advocate for CBT take the extant research results as strong encouragement and cheer themselves by the belief that more adequate tests will justify their confidence.

Theoretical Architecture

In addition to empirical support, theories are judged by their architecture. To fulfill ideal requirements, (a) the internal parts of a theory must fit together in a sensible way; (b) the theoretical argument must actually explain phenomena within its domain (that is, it must answer questions of “why” and “how”); (c) causal operations portrayed by a theory ultimately must be applicable to the real world; (d) the theory must accommodate regularities already established in the literature; and (e) it must solve, or at least address, long-standing empirical or intellectual problems for their users. Theoretical architecture is rarely discussed by scholars, and discussion, when stimulated, usually occurs because a theory being applied or tested is found wanting. For example, in attempting a test, a researcher may discover that a given theory contains a contradiction, which calls attention to a problem of architecture. Alternatively, consumers may find that when they attempt to derive policy applications, a given theory provides excellent descriptions of phenomena but fails to explain why those regularities exist. It is then that theoretical architecture looms large.

Yet, despite the obvious import of this criterion in judging theories, there seems to be no objective means for assessing it—certainly nothing comparable to the use of empirical evidence to test for adequacy. Therefore, in judging the architecture of CBT, the best that I can do is to reiterate how CBT presents itself, assuming that its features will speak for themselves.

Structure of Internal Parts

The model on which CBT is built resembles a wagon wheel. The hub represents the central causal process and is constantly revolving as people continually try to extend their control or avoid the consequences of others trying to extend theirs and as control-balancing transpires with feedback effects. This is the causal process previously described. The hub features extended spokes consisting of other extant theories and explanations that are feeding into the central causal process of CBT, providing contingencies such as self-control (Gottfredson & Hirschi, 1990) and morality (Wikström, 2006) or contributing elements of the causal structure itself (Agnew, 1992). Linkages of the outer parts of the spokes represent interconnections among theories and explanations external to the interior of the CB causal process.


The theoretical arguments that are draped on the frame of the wagon wheel explain deviant behavior; that is, they tell consumers why deviant behavior occurs. As a brief review, the “why” behind any deviance is to extend a social entity’s control, most often in response to a provocation that acutely reminds those with small CRs of that fact or stimulates in those with high CRs a sense that their high positions are not being honored. Provocations are often emotionally gripping, sometimes denigrating, and from time to time, humiliating and rage inducing. Their practical effect is to induce the person or other social entity experiencing them to contemplate deviance as a method for changing the control circumstances.

Contemplation, though usually habitual, rapid, and unconsciously undertaken, involves weighing the gain in control from deviance against the potential loss of control from reactions to that deviance. Control balancing often leads to the selection of some form of deviance reflecting both opportunity and minimal countercontrol. This is the “how” of deviance suggested by CBT. In addition, though, specific deviant acts themselves can be arrayed on a continuum on which the potential effectiveness of the act in extending one’s control and in avoiding direct, personal involvement converge. The specific range on the CBD continuum of acts that a person ends up committing can be predicted from the CR and other control-balancing variables. However, the entire process is subject to numerous contingencies that may influence how faithfully it unfolds.


Good architecture mandates that theories, which are basically coherently arranged intellectual ideas and frames, ultimately must be applicable to, and accurately reflect, empirical reality. In describing CBT, I have used many examples to show that the theory is rooted in real-world phenomena. But because the provided illustrations are hypothetical—that is, they are not based on specific instances—I will very briefly suggest three ongoing behaviors familiar to everybody in contemporary times for which CBT can potentially provide explanation.

One concerns the phenomenon of road rage, usually defined as actions—often way out of proportion to the perceived offense—stemming from emotions or feelings that sometimes come into play when an individual is offended by real or imagined misconduct by someone in or on a vehicle. I suggest that “road ragers,” if we were to investigate with CBT in mind, would turn out to be people with control imbalances, either low or high (but more likely low because road rage–induced actions appear to be of low CBD), who are often reminded of it (provoked). Also, we would likely learn that those who express road rage have previously been provoked many times in other contexts, and in response had tried to increase their control through numerous means, such as cutting in line at government offices, parking in disabled spaces without a permit, or stomping on improperly parked bicycles. Furthermore, at the time of the deviant act, these individuals most likely thought about potential countercontrols but perceived that trying to run the offending (or some other) vehicle off the highway, shooting a gun at the wayward vehicle or driver (or someone else), or yelling obscenities made sense because the perceived gain in control was worth the risk of the countercontrols that it would elicit.

A second form of familiar contemporary deviance is that of terrorism, one manifestation of which is attacking crowds of people with deadly weapons such as guns, vehicles, knives, or explosives, presumably for political purposes. Again, if such terrorist acts were studied with CBT in mind, one should find that terrorists are people with control imbalances, either for themselves or symbolically for the countries or religious groups of which they are a part. Moreover, it may be that the specific provocations that activate terrorist acts long in readiness are as likely to be simple, everyday slights as to be stinging, derogatory political statements by high government officials. Further, control balancing in the case of religiously based terrorism is likely to draw on perceptions of other worldly controls as being of more import than earthly ones. Finally, those who reside in the country where the terror is committed would probably have long histories of doing deviance much less drastic than the terrorist act because they face multiple and repeated provocations made more salient by peculiar modes of thought.

A third familiar deviant activity is misconduct by airline passengers and personnel. Focusing here on passengers, it seems that rowdiness is most likely to occur in the coach section of airplanes that also have first class compartments, especially when coach passengers are required to walk through the first class area to access their seats in the rear (DeCellas & Norton, 2017). One can imagine that many people in economy class experience situational low CRs. As passengers, they are controlled thoroughly from beginning to end, including enduring body and luggage searches, paying for transporting bags, waiting long periods of time for late-arriving flights, being herded onto and off planes, being restricted in seat and foot space, enduring sometimes too cold and sometimes too hot conditions, going hungry, and from time to time being rudely treated by airline employees. Moreover, normal sources of control that might put them in a more balanced situation are suspended during air travel. It would be surprising if directly passing through a compartment full of early-boarding, first class passengers enjoying enlarged space and being served drinks and food by flight attendants did not inspire considerable awareness of subordination among coach passengers, provoking some to go through the control-balancing process and end up committing some form of deviance.

Accommodation to the Issues

In the introduction to this article, I identified several long-standing concerns for social scientists, including cross-level transferability (micro and macro) and explanations, theoretical integration, SES salience in explanation of deviant conduct, and simultaneous consideration of past, present, and situational influences. I tried to show that CBT offers a response to all four of these concerns. Yet, good architecture also requires that theories provide believable explanations for the known facts about deviance.

Unfortunately, there appear to be few compelling regularities in the world of deviant behavior (Antonaccio, Tittle, Botchkovar, & Kranidioti, 2010; Delisi & Vaughn, 2016) to serve as tests of the architectural adequacy of general theories. While the basic patterns of negative associations between deviance proneness and age and female gender remain, those relationships are not as strong as once believed. And, as noted before, the SES pattern lacks reliability. Nevertheless, as shown in detail in the original statement of the theory (Tittle, 1995, Ch. 9), CBT not only accommodates assumed patterns, but it also helps explain deviations from those basic patterns.


In this article, I have tried to assess the standing of CBT in terms normally used to evaluate social science theories—empirical adequacy and the architecture of the theory. Adequacy of CBT in generating accurate empirical predictions is found to be mixed, although there is reason to believe that the testing implemented so far has shown an inadequate appreciation of important features of the theory. Nevertheless, enough empirical success has been reported to pose tantalizing possibilities for the future. Meanwhile, the architecture of the theory has undergone some modification in response to criticism (Braithwaite, 1997; Jensen, 1999; Savelsberg, 1999), but now it seems to be somewhat exemplary as a model for theoretical advancement.

Suggested Readings

Birkbeck, C., & LaFree, G. L. (1993). The situational analysis of crime and deviance. Annual Review of Sociology, 19, 113–137.Find this resource:

Gibbs, J. P. (1989). Control: Sociology’s Central notion. Urbana: University of Illinois Press.Find this resource:

Tittle, C. R. (1997). Thoughts stimulated by Braithwaite’s analysis of Control Balance theory. Theoretical Criminology, 1, 99–110.Find this resource:

Tittle, C. R. (1999). Continuing the discussion of control balance. Theoretical Criminology, 3, 344–352.Find this resource:

Tittle, C. R. (2000). The causal process of control balance theory. In R. D. Crutchfield, G. S. Bridges, J. G. Weis, & C. Kubring (Eds.), Crime readings (2d ed., pp. 408–420). Newbury Park, CA: Pine Forge Press.Find this resource:

Tittle, C. R. (2001). Control balance. In R. Paternoster & R. Bachman (Eds.), Contemporary theories (pp. 315–334). Los Angeles: Roxbury.Find this resource:

Tittle, C. R. (2004a). Control balance theory and violence. In H. Brownstein, M. A. Zahn, & S. L. Jackson (Eds.), Violence: From theory to research (pp. 51–69). Newark, NJ: Lexus/Nexus/Anderson.Find this resource:

Tittle, C. R. (2016). Introduction: Theory and contemporary criminology. In A. R. Piquero (Ed.), Handbook of criminological theory (pp. 1–17). West Sussex, UK: Wiley Blackwell.Find this resource:

Tittle, C. R., & Paternoster, R. (2000). Social deviance and crime: An organizational and theoretical approach. Los Angeles: Roxbury Press.Find this resource:


Agnew, R. (1992). Foundation for a general strain theory of crime and delinquency. Criminology, 30, 47–87.Find this resource:

Agnew, R. (2007). Pressured into crime: An overview of general strain theory. New York: Oxford University Press.Find this resource:

Agnew, R. (2016). Strain, economic status, and crime. In A. R. Piquero (Ed.), Handbook of criminological theory (pp. 209–229). West Sussex, UK: John Wiley and Sons.Find this resource:

Antonaccio, O., Tittle, C. R., Botchovar, E., & Kranidioti, M. (2010). The correlates of crime and deviance: Additional evidence. Journal of Research in Crime and Delinquency, 47, 297–328.Find this resource:

Baron, S. W., & Forde, D. R. (2007). Street youth crime: A test of control balance theory. Justice Quarterly, 24, 335–355.Find this resource:

Baumer, E. P., & Arnio, A. N. (2016). Macro-level theory: A critical component of criminological exploration. In A. R. Piquero (Ed.), Handbook of Criminological Theory (pp. 445–474). West Sussex, UK: John Wiley and Sons.Find this resource:

Braithwaite, J. (1981). The myth of social class and criminality reconsidered. American Sociological Review, 46, 36–57.Find this resource:

Braithwaite, J. (1997). Charles Tittle’s control balance and criminological theory. Theoretical Criminology, 1, 77–97.Find this resource:

Brauer, J. R., & Tittle, C. R. (2017). When crime is not an option: Inspecting the moral filtering of criminal actions alternatives. Justice Quarterly, 34, 818–846.Find this resource:

Burger, J. M. (1992). Desire for control: Personality, social, and clinical perspectives. New York: Springer.Find this resource:

Castro, E. D., Nobles, M. R., & Zavala, E. (2017). Assessing intimate partner violence in a control balance theory framework. Journal of Interpersonal Violence.Find this resource:

Curry, T. R. (2005). Integrating motivating and constraining forces in deviance causation: A test of causal chain hypotheses in control balance theory. Deviant Behavior, 26, 571–599.Find this resource:

Curry, T. R., & Piquero, A. R. (2003). Control ratios and defiant acts of deviance: Assessing additive and conditional effects with constraints and impulsivity. Sociological Perspectives, 46, 397–415.Find this resource:

DeCellas, K. A., & Norton, M. L. (2017). Physical and situational inequality on airplanes predicts air rage. Proceedings of the National Academy of Science, 113, 5588–5591.Find this resource:

Delisi, M., & Hochstetler, A. L. (2002). An exploratory assessment of Tittle’s control balance theory: Results from the National Youth Survey. Justice Professional, 15, 261–272.Find this resource:

Delisi, M., & Vaughn, M. G. (2016). Correlates of crime. In A. R. Piquero (Ed.), Handbook of criminological theory (pp. 18–36). West Sussex, UK: John Wiley and Sons.Find this resource:

Dunaway, R. G., Cullen F. T., Piquero, A. R., Wood, P. B., Burton, V. S., Jr., & Evans, T. D. (1999). Job autonomy and work place crime: A test of control balance theory. Paper presented at the Annual Meeting of the American Sociological Association, Toronto, November.Find this resource:

Fox, K. A., Nobles, M. R., & Lane, J. (2016). Control balance behind bars: Testing the generality of Tittle’s theory among incarcerated men and women. Crime and Delinquency, 62, 925–953.Find this resource:

Gottfredson, M., & Hirschi, T. (1990). A general theory of crime. Stanford, CA: Stanford University Press.Find this resource:

Hickman, M., & Piquero, A. R. (2001). Exploring the relationships between gender, control balance, and deviance. Deviant Behavior, 22, 323–351.Find this resource:

Hickman, M., Piquero, A. R., Lawton, B. A., & Greene, J. R. (2001). Applying Tittle’s control balance theory to police deviance. Policing: An International Journal of Police Strategies and Management, 24, 497–519.Find this resource:

Higgins, G. E., & Lauterbach, C. (2004). Control balance theory and exploitation: An examination of contingencies. Criminal Justice Studies, 38, 241–260.Find this resource:

Higgins, G. E., Lauterbach, C., & Tewksberry, R. (2005). Control balance theory and violence: An examination of contingencies. Sociological Focus, 38, 241–260.Find this resource:

Hughes, L. A., Antonaccio, O., & Botchkovar, E. V. (2015). How general is control balance theory? Evidence from Ukraine. Justice Quarterly, 32, 950–975.Find this resource:

Jensen, G. A. (1999). A critique of control balance theory: Digging into the details. Theoretical Criminology, 3, 339–343.Find this resource:

Messner, S. F., Krohn, M. D., & Liska, A. E. (Eds.). (1985). Deviance and crime: Problems and prospects. Albany: State University of New York Press.Find this resource:

Nobles, M. R., & Fox, K. A. (2013). Assessing stalking behaviors in a control balance theory framework. Criminal Justice and Behavior, 40, 763–783.Find this resource:

Piquero, A. R., & Hickman, M. (1999). An empirical test of Tittle’s control balance theory. Criminology, 44, 319–341.Find this resource:

Piquero, A. R. (2002). The rational choice implications of Control Balance Theory. In A. R. Piquero & S. G. Tibbetts (Eds.), Rational choice and criminal behavior: Recent changes and future challenges (pp. 85–107). New York: Routledge.Find this resource:

Piquero, A. R. (2003). Extending Tittle’s control balance theory to account for victimization. Criminal Justice and Behavior, 30, 282–301.Find this resource:

Piquero, A. R., & Piquero, N. L. (2006). Control balance and exploitative corporate crime. Criminology, 44, 397–430.Find this resource:

Rafeld, H., Fritz-Morgenthal, S., & Posch, P. (2017). Behavioral patterns in rogue trading. Analysing the cases of Nick Leeson, Jerome Kerviel, and Kweku Adoboli in light of the control balance theory. Journal of Financial Compliance, 1(2–3).Find this resource:

Savelsberg, J. J. (1999). Human nature and social control in complex society: A critique of Charles Tittle’s control balance. Theoretical Criminology, 3, 331–338.Find this resource:

Short, J. F., Jr. (1998). The level of explanation problem revisited—the American Society of Criminology 1997 Presidential Address. Criminology, 36, 3–36.Find this resource:

Tittle, C. R. (1983). Social class and criminal behavior: A critique of the theoretical foundation. Social Forces, 61, 334–358.Find this resource:

Tittle, C. R. (1995). Control balance: Toward a general theory of deviance. Boulder, CO: Westview Press.Find this resource:

Tittle, C. R. (2004b). Refining control balance theory. Theoretical Criminology, 8, 395–428.Find this resource:

Tittle, C. R. (2010). Control balance theory and social policy. In H. Barlow & S. Decker (Eds.), Criminology and public policy: Putting theory to work (pp. 6–24). Philadelphia: Temple University Press.Find this resource:

Tittle, C. R., Villemez, W., & Smith, D. A. (1978). The myth of social class and criminality: An empirical assessment of the empirical evidence. American Sociological Review, 43, 643–656.Find this resource:

Tittle, C. R., Villemez, W., & Smith, D. A. (1982). One step forward, two steps back: More on the class/criminality controversy. American Sociological Review, 47, 435–438.Find this resource:

Wikström, P.-O. H. (2006). Individuals, settings, and acts of crime: Situational mechanisms and the explanation of crime. In P-O. H. Wikstrom & R. J. Sampson (Eds.), The explanation of crime: Context, mechanisms, and development (pp. 61–107). Cambridge, UK: Cambridge University Press.Find this resource:

Wikström, P.-O. H., & Treiber, K. (2016). Situational theory: The importance of interactions and action mechanisms in the explanation of crime. In A. R. Piquero (Ed.), Handbook of Criminological Theory (pp. 415–444). West Sussex, UK: John Wiley and Sons.Find this resource:

Williams, K. S. (2008). Using Tittle’s control balance theory to understand computer crime and deviance. International Review of Law, Computers, and Technology, 22, 145–155.Find this resource:

Wood, P. B., & Dunaway, R. G. (1997–1998). An application of control balance theory to incarcerated sex offenders. Journal of the Oklahoma Criminal Justice Research Consortium, 4, 1–12.Find this resource: