Street Youth Crime: A Test of Control Balance Theory

Written By: Sam Savage

Published Date: June 28, 2007 Last Edited: June 28, 2007

By Baron, Stephen W Forde, David R

Utilizing a sample of 400 homeless street youths, the paper explores the role control balance plays in the generation of crime. Using vignettes designed to represent violent crime, serious property crime, and minor property crime, the paper tests whether these youths sense of control over their poverty, shelter, hunger and other living conditions influences their participation in crime. Further, it examines how perceptions of risk and thrill, as well as deviant values, self-control, deviant histories, and peer support impact on crime. Results indicate that both control deficits and control surpluses were related to assault and serious theft but not minor theft. Perceptions of thrill, deviant peers, deviant histories, and deviant values predicted violent and property crime, and perceptions of risk were related to the property offenses. Criminal peers also conditioned the impact of control surpluses and deficits on property offenses. Results are discussed in terms of future research and policy. Keywords street youth; control balance theory; violent crime; property crime

Introduction

In 1995, Charles Tittle published his award-winning monograph, Control Balance. Since publication, control balance theory has been reviewed, critiqued, empirically tested, and refined (Tittle, 2004). In this article, we briefly review the control balance explanation of deviance and contribute to the emerging literature on control balance theory by presenting an empirical test using a sample of homeless youth.

In control balance theory, Tittle suggests that behavioral restrictions and deviant motivations need to be examined in tandem to understand deviance. Central to his argument is the “control ratio” defined as “the total amount of control to which an individual is subject relative to the total amount of control that he or she can exercise” (Tittle, 1995, p. 148). Being controlled reflects limitations placed on individuals by various physical and social arrangements when attempting to satisfy goals and motivations, while exercising control reflects the extent to which an individual can escape these limitations or help or hinder others in their pursuits. People are said to have a number of specific control ratios associated with their various roles, statuses, and environments, as well as a general control ratio reflecting their overall ability to control and be controlled (Tittle, 2001, p. 317).

When the amount of control a person exercises is equal to the amount of control they are subject to, their control ratio is balanced, leaving them less likely to be involved in deviance. In contrast, control ratio imbalances, which can take two forms, leave one more likely to consider deviance as a method of altering the control ratio. “Control surpluses” emerge where the control exercised is greater than the control subject to. “Control deficits” arise where the control subject to exceeds the control exercised. While imbalances predispose a person to undertake deviance to improve control ratios, the motivation for deviance only emerges under circumstances where people experience a provocation that reminds them, or makes them aware, of their control imbalance creating negative emotions (Tittle, 2004, p. 411). The greater the imbalance, the higher the likelihood that people will feel debased or humiliated when faced with provocation. This focuses the individual on deviance as a potential solution and the benefits, rewards, and perceived pleasures associated with engaging in the deviant act that might serve to alter the control ratio (Piquero & Hickman, 2002, p. 86).

Motivation is only transformed into deviance if there is “opportunity.” The person must have both the capabilities and the occasion to undertake the particular behaviors being considered. Given motivation and opportunity, people will consider the deviant acts they perceive as leading to the greatest change in their control ratio. These behaviors, however, are also likely to be viewed as serious and bring about reactions of counter control (Tittle, 2001, p. 321). Therefore, the most serious acts may not serve as viable options for altering a control imbalance leaving individuals to choose other forms of deviance. The act actually chosen will be influenced by “constraint.”

Constraint is a complex variable that represents the seriousness of the act being considered and situational risk (Tittle, 2004, p. 414). Seriousness is related to the degree of counter control or the “amount of control inherent in an act … that could possibly be forthcoming” (Tittle, 2004, p. 403) while “situational risk” is the concern about the act being exposed and actually experiencing the counter controlling reactions. Thus, individuals consider constraint when reviewing possible avenues to address imbalances.

The causal process is said to work as outlined when a person has high selfcontrol. Low self-control makes people vulnerable to provocation leaving them more apt to become motivated for crime, more likely to react instantly, less likely to consider the counter control associated with their acts, and less likely to consider gains (Tittle, 2004, p. 416). In contrast, those high in selfcontrol can restrain themselves from taking instantaneous actions leaving them more likely to consider different types of acts than those with low selfcontrol, and these acts may have more impact on altering control balance ratios.

Types of Deviance

Given a favorable configuration of opportunity and constraint, an unbalanced control ratio will lead to deviance. Forms of deviance can be distinguished by two components of “control balance desirability” (Tittle, 2004, p. 405). The first surrounds the potential long-term change in the control ratio that may result from the act. The second focuses on the degree to which the offender needs to be personally or directly involved with the victim (Tittle, 2004, p. 405). Acts with little direct contact with victims that allow for long-term changes in control balance ratios have the greatest control balance desirability. Acts that involve direct contact, and have shorter-term impact on control balance ratios, are seen to have less control balance desirability. Tittle (2004, p. 406) suggests that acts can be assorted on a continuum based on these characteristics. He argues, however, that the theory does not permit the precise estimate of particular behaviors but claims it can help explain the choosing of a certain act from behaviors with similar control balance desirability scores (Tittle, 2004, p. 407). He also allows that including a greater number of the variables outlined in the theory beyond control ratios may improve the prediction of specific acts (Tittle, 2004, pp. 408, 415).

Actors with substantial control surpluses, high in self-control, who are presented with opportunities and few constraints will be most likely to undertake behaviors high in control balance desirability. In contrast, acts low in control balance desirability are more likely to be accessed by people who have small to medium control deficits, are low in self-control, face constraints on committing high control balance desirability acts, but have opportunities for the low desirability acts. Acts around the middle of the desirability continuum may point towards various levels and amalgamations of the causal variables. Those with surpluses or small deficits, a range of self-control, and a range of constraint could engage in acts of medium control balance desirability as long as opportunities present themselves. Those with extreme deficits, however, are most likely to resort to “submission” which is not on the continuum (see Tittle, 2004).

Contingencies

The causal process of the theory may be influenced by various “contingencies” including moral beliefs, self-efficacy, prior deviant experience, and subcultural involvement. While not required for the control balance variables to impact on deviance, and not having significant independent causal impact, contingencies can modify the control balancing process (Tittle, 1995, pp. 201, 214). For example, moral beliefs may disturb the impact of an imbalanced control ratio on crime (Tittle, 1995, pp. 208-209). In contrast, when facing control imbalance, a higher self-efficacy may leave one more likely to confront controls and/or take advantage of deviant opportunities others would find risky. Prior experience with deviance can also act as a contingency because people can recall previous successes in altering control imbalances through crime and recognize the low potential for counter control (Tittle, 1995, p. 216). Similarly, subcultural involvement may produce motivations for deviance beyond those stemming from imbalanced control ratios as people are exposed to various forms, excuses, rationalizations, and expectations of deviance (Tittle, 1995, p. 220).

Risk, an important component of constraint, can also act as contingency. The link between control imbalance and deviance should be greatest when actors perceive the risk of detection and punishment to be minimal. While admitting that Tittle does not make this explicit, Piquero and Hickman (2002, p. 86) argue that benefits or perceived rewards represented in the motivation variable can also be used as a contingency variable. Here, the greater the benefits, including the perceived pleasure of the act, the greater the impact of an imbalanced control ratio on crime. Tittle (1995) also dedicates some detail to the link between socioeconomic status and deviance through its influence on control ratios, motivation, constraint, contingencies, and opportunities. Control surpluses, the desire to increase surpluses, and opportunities, lead people in higher socioeconomic locations to engage in “elite” deviance although the behavior of people in these positions is still restricted by various social, political, and legal constraints, and the majority may have balanced control ratios (Tittle, 1995, p. 257). In contrast, lower-status people tend to be deficient in various political and financial resources that might enable them to control or advance their situations, and they are subject to intense surveillance and repression by state authorities (Tittle, 1995, p. 258). This argument suggests the greater likelihood of control deficits, although subcultural resources in lower-class areas can enable people to balance control ratios.

While the contingencies for deviance will not vary substantially by socioeconomic status, those in lower socioeconomic locations may be more involved in “street” types of offenses because of their criminal subcultural participation and prior experience in these types of crimes. In contrast, those in higher socioeconomic locations should have greater experience in “elite” forms of deviance and be exposed to models and expectations that support these behaviors. Tittle (1999, p. 259) concedes that lacking “precise information” predicting deviance for different economic locations is difficult.

Past Research

Control balance theory has been subject to a limited number of empirical tests. The bulk of this research finds both control balance deficits and control balance surpluses to be related to deviance including assault (Piquero a Hickman, 1999, 2002), drug and alcohol use (Curry & Piquero, 2003), deviant sexual practices (Piquero & Hickman, 1999), eating disorders (Hickman & Piquero, 2001) using others’ school work (Hickman & Piquero, 2001; Higgins & Lauterbach, 2004), and cheating (Curry, 2005). Further, in a theoretical extension, control imbalances were related to property and general victimization (Piquero & Hickman, 2003). Only one study has found that deficits alone were related to deviance (Hickman, Piquero, Lawton, & Greene, 2001).

The research has also examined the impact of other variables in the causal process. There is some support for the relationship between constraint and deviance (Curry, 2005; Curry ft Piquero, 2003; although see Higgins & Lauterbach, 2004). The effect of risk (one component of constraint) on deviance, however, has been mixed (Hickman & Piquero, 2001; Piquero & Hickman, 1999) as has been the support for the effect of low self-control (Curry, 2005; Hickman & Piquero, 2001; Higgins a Lauterbach, 2004; Piquero a Hickman, 1999, 2002; see also Curry a Piquero, 2003, for impulsivity). Work exploring the direct effects of the contingency variables shows moral beliefs to decrease the likelihood of deviance (Hickman a Piquero, 2001; Higgins a Lauterbach, 2004; Piquero a Hickman, 1999, 2002), while deviant histories have shown no direct impact on deviance (Hickman a Piquero, 2001; Higgins a Lauterbach, 2004; Piquero a Hickman, 1999, 2002).

The limited work exploring the conditioning effects of the contingency variables has found control surpluses and deficits to have a greater impact at higher levels of impulsivity (Curry a Piquero, 2003). The findings on the conditional impact of constraint on the control balance ratio have been mixed (Curry a Piquero, 2003; Higgins a Lauterbach, 2004). However, surpluses and deficits have been found to have a greater impact at lower levels of risk, and higher levels of pleasure or sensation seeking (Higgins a Lauterbach, 2004; Piquero a Hickman, 2002) as well as at levels of low risk and low pleasure, which is in contrast to expectations (Piquero a Hickman, 2002).

While informative, these works rely extensively on college populations and similar control balance items (see Curry, 2005; Curry a Piquero, 2003; Hickman a Piquero, 2001; Higgins a Lauterbach, 2004; Piquero a Hickman, 1999, 2002, 2003; Piquero et al., 2001). Only one study has used a noncollege population (Hickman et al., 2001). Thus, it is difficult to determine the general support for the theory. Most research has examined only the main effects of the contingency variables (although see Curry a Piquero, 2003; Higgins a Lauterbach, 2004; Piquero a Hickman, 2002), no research has examined the conditioning effect of deviant histories, deviant peers, self-efficacy, or moral beliefs, and none of this research has examined property offending.

To date, no research has utilized a street population, so it is unknown if the types of control deficits and surpluses these people experience are related to deviant behavior. In light of Tittle’s discussion of SES, it seems important to explore how control balance ratios, and the various variables that impinge on these ratios, influence the deviance of lower socioeconomic individuals. We utilize a sample of homeless street youth to explore control balance theory. This is a population with serious and persistent problems of crime and adversity (Hagan & McCarthy, 1997a). They are faced with crises of day-to-day survival, including finding food, shelter, and employment that might lead to control balance deficits. Research suggests they hold values supportive of crime, are involved in criminal networks, and have deviant histories (Baron, Kennedy, & Forde, 2001; Hagan & McCarthy, 1997b), factors that might condition the impact of control imbalances. Further, evidence suggests that perceptions of risk (Baron & Kennedy, 1998) and personality factors (Baron, 2003) influence their criminal behavior.

Method

Researchers define street youth as young people who have runaway or been “thrown” from their homes and/or who spend much of their time in public areas. Street youth are a heterogeneous group that can include students, dropouts, the employed, and the unemployed, who may or may not have homes to return to at night (see Shane 1996; Whitbeck a Hoyt, 1999). This study utilized 400 respondents (265 male, 135 female) who were identified based on four sampling criteria: (1) to cover the age range of those described as street youth participants must be aged 24 and under; (2) they must have left or finished school; (3) they must be currently unemployed; (4) they have spent time without a fixed address or living in a shelter in the previous 12 months. These criteria were developed (1) to exclude those in school and eliminate those not eligible for full- time employment; and (2) to obtain a sample of “serious””at risk” youth.

Data Collection

In-depth interviews were completed between May 2000 and August 2001 in Vancouver, Canada. The study took place in and around the city’s downtown business core bordered by the local skid row and “inner city.” The area contained a mix of commercial and financial establishments surrounded by bars, pawnshops, sex shops, and tattoo parlors, as well as potential residences for the study sample including single-room-occupancy hotels, shelters, rundown residential units, and abandoned buildings.

Potential respondents were approached by one of the researchers, alerted to the project and screened for eligibility. Youths who met the selection criteria were provided additional information and solicited to participate. Interested youths were presented with informed consent forms outlining study goals and their rights within the interview. Youths granting consent were then interviewed in locations that allowed for some privacy as well as familiarity including fast food restaurants, parks, in front of store-front social services, bus shelters, and on the street. Interviews averaged an hour and 10 minutes in length, and respondents were awarded $20 in food coupons at a popular fast-food restaurant. Other contacts were initiated by youths who learned of the project and solicited interviews or through introductions from previously interviewed respondents.

The 400 youths interviewed had an average age of almost 20 years (O = 19.9). The racial make-up of the sample was predominately Caucasian (83 percent). Aboriginal youths made up the majority of the other respondents (12 percent).1 The average length of homelessness in the previous 12 months was close to 7 months (M = 6.8).

Scenarios

Each respondent was read three scenarios representing three different forms of deviance and asked to respond to a series of questions regarding each scenario. Although this strategy has been used in most of the empirical research on control balance theory (Curry, 2005; Hickman St Piquero, 2001; Hickman et al., 2001; Higgins a Lauterbach, 2004; Piquero a Hickman, 1999, 2002, 2003; although see Curry a Piquero, 2003), it has been critiqued because the intention to offend does not necessarily translate into actual offending (Piquero a Hickman, 1999). However, research exploring this issue, including meta-analyses, shows a high correlation between intentions and actual behavior (Fishbein a Ajzen, 1975; Green, 1989; Kirn a Hunter, 1993; Piquero a Hickman, 1999). The following is an example of the scenario involving assault.

Brian is standing on a street corner drinking a pop. A guy named David is rushing to catch the bus and bumps into Brian just enough to cause him to spill the pop all over the sidewalk. Brian’s friend’s can’t help but laugh at him. David stops and says “Sorry man.” As David turns to continue for the bus Brian grabs his shoulder and says “You spilt my pop you asshole. Get me another one.” David tells him to “fuck off.” Brian punches him in the face and a fight breaks out between them.

After each scenario was read to the respondent they were asked on a scale from zero (no chance at all) to 10 (100 percent chance) to estimate the probability that they would do what the person in the scenario did (in the case of the assault scenario example Brian). We use two other scenarios (see Appendix A) to explore minor theft and serious theft (see Table 1 for descriptive statistics). Tittle (2004, p. 212) argues that when examining specific acts, researchers should explore their “probability” within a certain scope of control desirability. The three dependent variables were recoded to O or 1, where O indicated no probability of offending, and 1 indicated a nonzero probability of engaging in the scenario-specific form of deviance. We also report the results of the original scale in the text. We do not have information regarding respondent’s perceptions of the control desirability of these offenses. However, all of these offenses share the requirement of direct contact with the target which Tittle suggests places them lower on the desirability continuum. Measurement of Variables

In his book, Tittle (1995, p. 267) suggests that the “general perceptual method” is one strategy that can be utilized to measure control balance ratios. Like the empirical research conducted so far (see Curry, 2005; Curry & Piquero, 2003; Hickman & Piquero, 2001; Hickman et al., 2001; Higgins a Lauterbach, 2004; Piquero & Hickman 1999, 2002, 2003), we ask respondents to estimate the amount of control they exercise over certain circumstances and then ask them to estimate the amount of control these circumstances have over them. Piquero and Hickman (1999) stress that different samples will have different life domains, and these must be considered in the measuring of a control balance ratio. With this in mind, individuals were asked to rate the degree of control they had over their homelessness, unemployment, regularity of food, avoiding physical illness, cleanliness, and being uncomfortable because of the cold. Response options ranged from O (no control) to 10 (total control). Respondents were then asked to rate on a 0-10 scale the degree of control these same things had over them. Theoretically, each domain functions in the situationally specific circumstance as well as in a global context. However, Tittle (1995, p. 267) argues that the “the most important measure is of the overall general control ratio” which should be created from “specifically focused ratios.”

Following Tittle (1995, p. 267), the control balance ratio was created by first summing the items measuring the amount of control one is subject to. second, the items for the amount of control one can exercise were summed. A principal axis factor analysis was then undertaken to confirm that the two scales were unidimensional, and tests were conducted that indicated that the scales were reliable (control subject to, a = .741; control exercised, a = .737). Finally, to create an overall control balance ratio, we calculate the ratio of the amount of control to which one is subject relative to the amount of control one can exercise. In this scale, subjects who score over 1 have a control deficit (the larger the score over one, the larger the deficit), while following past research (see Piquero & Hickmam, 1999) those scoring equal to or less than 1 have a control surplus (the smaller the number under 1, the greater the surplus).2 The control balance ratio was then segmented to create a separate variable for a control balance surplus and a control deficit. Those with extreme deficits, those who reported having no control at all across the six domains (n = 3), were left out since theoretically they are linked to submission.3

Tittle (1995, pp. 167-168) argues that one of the other vital components in the control balance process is constraint, which is made up of the control balance ratio, perceptual risk, and seriousness. Here we examine perceptions of risk. Respondents were asked after each scenario “On a scale from O (not likely) to 10 (very likely) to rate the risk of getting caught by committing the act portrayed in the scenario.” One weakness of this measure is that it is not specific in who will discover the act. Following Piquero and Hickman’s (2002) argument for the examination of benefits, we explore if respondents thought they would get a sense of thrill engaging in the scenario presented. They were asked “On a scale of O (not exciting) to 10 (very exciting), how exciting would it be to engage in the act portrayed in the scenario?” To examine deviant values, respondents were asked after each scenario “On a scale of O (very wrong) to 10 (not wrong at all) how morally wrong is the act portrayed in the scenario?” To explore for the influence of peers, respondents were asked after each scenario “On a scale of O (none) to 10 (most) how many of your friends have engaged in the act portrayed in the scenario?” Self-efficacy is measured by the question “I generally have confidence that when I have a plan I will be able to carry it out” (1 = strongly disagree; 4 = strongly agree). As one of the few measures that is not scenario-specific, caution should be exercised since it is a single-item measure. Prior deviance was obtained by asking the respondents how many times in the past year they had done the following: broken into a car, broken into a building, taken something worth less than $50, taken something worth more than $50, broken into a structure to sleep, stolen food, taken a car without permission of the owner, used physical force to get money or things from another person, attacked someone with a weapon or fists injuring them so badly they probably needed a doctor, got into a fight, and taken part in a group fight. The raw scores of individual offenses were aggregated across the range of offenses to create indices of property and violent histories. To reduce skewness, these scales were logged and scores two standard deviations above the mean receded to that score. Consistent with the past research on control balance, the Grasmick, Tittle, Bursick and Arneklev (1993) scale is used to measure selfcontrol (a = .777). A principal axis factor analysis revealed that this scale broke into seven factors. However, an examination of the Scree plot to look for “discontinuity” in eigenvalue magnitudes indicated that the scale was unidimensional (see Arneklev, Grasmick, Tittle, & Bursick, 1993, p. 232, for a discussion). The 23 items drawn from this scale were summed and divided by their number. Finally, chronological age and gender are also controlled for in the analysis (see Table 1 for descriptive statistics and Appendix B for correlation matrix).

Nonlinear regression is used to analyze the models, since Tittle (1995, p. 176) argues that a number of the causal linkages are nonlinear. Our analytic strategy replicates the analytic methods used by Piquero and his colleagues in their previous empirical tests of control balance theory. Specifically, the equation utilized takes the form:

…

Y is the dependent variable, and CBR is the control ratio. This variable is segmented where values of 1 or less than 1 are defined as control balance surpluses, and those greater than 1 are defined as control balance deficits. The equation also assesses low self- control (lsc), the chances of getting caught (risk), the excitement of undertaking the act (thrill), how wrong the act is (deviant values), the degree of control over the situation (self-efficacy), the total amount of the crime in the previous 12 months (deviant experience), the number of peers who have undertaken the criminal act (peers), the age of the respondent (age), and their gender (males = O, females = 1).

Findings

Table 1 shows that the majority of youth in the sample had control surpluses (67 percent) with only 31 percent reporting deficits. Moving to Table 2, we see that having either a control deficit or a control surplus is a significant predictor of assault. The greater the deficit, the more likely street youth indicate that they would assault the person in the scenario. Respondents with smaller surpluses, in contrast, are more likely to report that they would engage in violence. In the analysis utilizing the original 0- 10 scale, neither control surpluses nor deficits predicted assault. Table 2 also reveals that perceptions of thrill increase the likelihood of reporting an intention to engage in assault. Perceptions of risk, in contrast, are not related to the intention to assault. Deviant peers, prior violent crime, being male, and deviant values are also related to intention to assault. Age, self- control, and self-efficacy are not significant predictors of intention to assault.

Table 2 indicates that neither control deficits nor control surpluses predict minor theft. This finding is replicated using the 0-10 scale for a dependent variable. In contrast, thrill, lack of risk, deviant peers, deviant values, and low selfcontrol are related to intentions to engage in minor theft. The final model in Table 2 shows that both control deficits and control surpluses are related to serious theft. Neither of the control imbalances were related to serious theft using the 0-10 scaling of the dependent variable. Further, thrill, risk, deviant peers, deviant values, and prior participation in property crime are also related to serious theft, as are being young and male.

Although Tittle argues that the control ratio will have a strong influence on the probability and type of deviance, he also suggests that this effect may not occur with the same strength or likelihood in all circumstances. Contingencies may alter, or impinge on, the operation of the control balancing process (Tittle, 1995, p. 201). Thus, the strength of the relationship between control imbalances and crime may be different, depending on levels of risk, subcultural participation, morality, self-control, self-efficacy and deviant histories. To explore this, the sample was split at the median for each of the contingency variables: risk, thrill, morality, self- control, self-efficacy, peers, and deviant history. The nonlinear regressions were repeated on each of the samples representing lower levels and higher levels of each specific variable. This allowed us to explore if the effects of control surpluses and control deficits on our dependent variables were different depending on the level of the contingency variables (see Appendixes C, D, and E). This process revealed the conditioning impact of deviant peers on control imbalances. At higher levels of deviant peers, control surpluses and control deficits had stronger relationships with both minor theft and serious theft. Again, these results held only for the dichotomized dependent variables. Discussion

This study set out to examine the role control balance ratios play in the generation of street youth crime. The results indicate that both control deficits and control surpluses were related to their violent and serious theft, and when deviant peers were available, these imbalances increased minor theft. It may be reasonable to suggest that for those with deficits, there is a sense of desperation to ensure that control is not decreased even further. As control deficits become larger, there is less for street youth to lose by engaging in certain types of criminal behavior. Or it may be that as their deficits grow, options involving offenses of higher control desirability decline. Or it could be a combination of these types of explanations. The findings also suggest that those with small control surpluses use assault, and serious theft, to extend their surpluses in situations where these meager surpluses are threatened. Lacking the surplus that would allow them to undertake offenses higher in control balance desirability, these youth undertake offenses that involve face-to-face contact.

The findings that both surpluses and deficits are related to the same behavior is consistent with past research. Further, the direction of the deficits and surpluses is also consistent with past work (although see Curry, 2005; Higgen & Lauterbach, 2004). The recently revised theory allows for both deficits and surpluses to explain similar behaviors that fall in the middle range of the control desirability continuum. Those with deficits who engage in middle-range acts are expected, however, to have smaller deficits. The findings here then raise some questions. Since these behaviors are more likely to be undertaken as deficits increase and surpluses decrease, there needs to be some direction regarding what should be considered minor or moderate deficits and surpluses. As observed by others, Tittle provides little guidance regarding the “relative positions on the control balance continuum” (Piquero & Hickman, 1999, p. 337), and the recent refinement does little to address this question.

Further, the findings suggest that more work is necessary to make clear where on the control desirability continuum the offenses examined here fit. If both surpluses and deficits are related to offenses towards the middle of the control balance continuum, then serious theft and assault, despite requiring face-to-face contact with targets, may have more than a short-term impact on the control balance ratio for a population living in extreme poverty. In contrast, the minor-theft offense, which shares the face-to-face component, was only influenced by control imbalances under conditions where there was deviant peer association. This suggests that minor theft provides at best only short-term alteration of control ratios and only under restricted circumstances. The finding that deficits and surpluses are related to assault suggests that types of violent offenses may contain more than just high “personal desirability” (see Tittle, 2004). This is consistent with the research in college populations, and it reveals that the population being examined may also be important in determining the context of control desirability. So while Tittle (1995) argues that those in various socioeconomic locations may have equal opportunity to engage in the offenses examined here, it may be that the differential participation in these activities is based on the different appeal in altering control balance ratios because the desirability varies across different social groups for similar offenses. At the same time, certain offenses may share appeal across social groups.

The findings also imply that control imbalances may be better predictors of any probability of offending rather than predictors of the probability of offending. This appears to be consistent with Tittle’s (2004) caution regarding the accuracy with which the theory can be used to predict behavior. At the same time, this would appear to be a drawback in that the theory cannot distinguish between those with weak and strong intentions leaving it perhaps to overpredict behavior.

Tittle has argued that the effect of imbalances will be influenced by various contingencies. There is only limited support for this argument here. In both cases, the impact of control deficits and control surpluses was greater under conditions where the respondents had deviant peers. Peers may help to contextualize experiences as imbalances, as well as offer and support crime as a method for altering control ratios. Generally, however, the impact of deficits and surpluses is not significantly different at varying levels of the contingency variables.

Instead, these contingency variables appear to have strong direct effects on crime. Deviant values had a direct impact on crime, and unlike past research, deviant histories were related to the intention to offend. Prior offending may educate street youth on the realistic risks and gains that can alter control balance ratios. Risk, one component of the constraint variable, was predictive of property but not violent offending. Perhaps the emotional nature of this offense overwhelms the consideration of risk. Further, thrill appears important in the generation of crime. Katz (1988) argues that crime has a seductive quality. Different crimes have different forms of seduction but they share the ability to provide excitement to people and at the same time provide relevance to the identity of the offender. Crimes are not only fun but also have a transformative quality about them that it can be argued alter the control balance ratio.

Tittle’s (2004) recent addition of self-control to the control balance process is not supported. It was not associated with any of the individual offenses where control imbalances predicted crime. Past work including low social control controlling for control imbalances has been mixed, but the work examining violent offenses also found no effect (Piquero 6t Hickman, 1999, 2002). Further, there is little support that control ratios are conditioned by self- control. This suggests that in this population, the causal process works the same way regardless of the level of self-control.

This study is important because Control Balance theory has yet to be applied to an at-risk population. The study allows for comparison with the results of research on more conventional populations and can establish the generalizability of the theory. The study examines unstudied offenses (serious and nonserious property crimes) and a range of offenses and tests for a number of the yet-to-beexplored contingencies. At the same time, caution must be exercised in that some of the measures are single-item; certain variables like a full measure of constraint, were not included; the method of sampling and the restricted sample make it hard to generalize the findings to other populations; and the use of scenarios captures only the intention to offend. Overall, this project is one step of many that will be required to explore this complex theory, and it is only through this type of process that control balance theory can be refined and its place as a general theory of deviance determined.

The findings also present a challenge to policymakers. Tittle (2001, p. 331) himself stresses that the theory was not developed with policy in mind and admits to avoiding such issues (see Tittle, 1997). The fact that both deficits and surpluses lead to the forms of crime makes the policy issue even more challenging. One might attempt to address deficits through employment, housing, and the provision of food, which in turn would decrease crime. These types of responses, however, may also be appropriate for those with surpluses. While it seems problematic to decrease the surpluses of those who feel they have some control over the domains examined here, it is also probably the case that their control over these domains is a reflection of their survival skills. They have learned how to get food, find alternative shelter, avoid the elements, and access alternative avenues for money. Substituting these with more conventional avenues might decrease their need to be resilient under harsh conditions, leaving them less likely to engage in crime. These types of changes, however, still pose challenges in terms of funding and politics. Tittle (2001 ) argues that one alternative may be to increase constraint. However, this is a population already high in constraint, and central to the theory is that control can cause crime. This suggests that increases in risk, and the degree of counter-control associated with offending could serve to increase crime in this population or lead to different types of offending. Perhaps a more productive avenue is to reduce provocation (Tittle, 2001). Perhaps the general public needs to be educated about certain issues including the causes of homelessness and the various challenges that the homeless face. Thus, policy might focus on not only trying to provide things such as housing and employment, but also encouraging compassion for this disadvantaged population.

Acknowledgments

The authors would like to acknowledge the financial support of the Social Sciences and Humanities Research Council of Canada and Queen’s University Chancellor’s Research Award. They would also like to thank Jennifer Robinson for her research assistance. They would also like to thank Chester Britt and the anonymous reviewers for their helpful comments. 1. Aboriginals are drastically overrepresented in the sample. According to Peters and Murphy (1993), only about 1 percent of the youths in city schools are native.

2. For example, if the respondent’s total control subject to was 30, and their control exercised was 15, their control balance ratio would be 2, which is a deficit. In contrast, if the respondent’s total control subject to was 15, and their control exercised was 30, their control balance ratio would be .5, which is a surplus.

3. Following Piquero et al.’s (2001) argument regarding the inefficiency of the control balance scale, we also followed their work and recoded the items to 4 points (O = O, 1-4 = 1, 5-8 = 2, 9- 10 = 3). The correlation between scale created with the recoded items and the original scale was .972. Since the results using either scale were similar, we retain the scale using the original coding.

References

Arneklev, B. J., Grasmick, H. G., Tittle, C. R., a Bursick, R. J. (1993). Low self-control and imprudent behavior. Journal of Quantitative Criminology, 9, 225-247.

Baron, S. W. (2003). Self control, social consequences, and criminal behavior: Street youth and the general theory of crime. Journal of Research in Crime and Delinquency, 40, 403-425.

Baron, S. W., & Kennedy, L. W. (1998). Deterring crime among street kids: An empirical test. Canadian Journal of Criminology, 40, 27-60.

Baron, S. W., Kennedy, L. W., & Forde, D. R. (2001). Male street youths’ conflict. The role of background, subcultural and situational factors. Justice Quarterly, 18, 759-789.

Curry, T., & and 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.

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.

Fishbein, M. M., & Ajzen, I. L. (1975). Belief, attitudes, intention and behavior. Reading, MA: Addison-Wesley.

Grasmick, H. G., Tittle, C. R., Bursick, R. J., a Arneklev, B. J. (1993). Testing the core empirical implications of Gottfredson and Hirschi’s general theory of crime. Journal of Research in Crime and Delinquency, 30, 5-29.

Green, D. (1989). Measures of illegal behavior in individual- level deterrence research. Journal of Research in Crime and Delinquency, 26, 253-275.

Hagan, J., & McCarthy, B. (1997a). Anomie, social capital, and street criminology. In N. Passas & R. Agnew (Eds.), The future of anomie theory (pp. 121-141). Boston: Northeastern University Press.

Hagan, J., & McCarthy, B. (1997b). Mean streets: Youth crime and homelessness. Cambridge, MA: Cambridge University Press.

Hickman, M., & Piquero, A. (2001). Exploring the relationships between gender, control balance and deviance. Deviant Behavior, 22, 323-351.

Hickman, M. J., Piquero, A. R., Lawton, B. A., a Greene, J. R. (2001). Applying Tittle’s control balance theory to police deviance. Policing, 24, 497-519.

Higgins, G. E., a Lauterbach, C. (2004). Control balance theory and exploitation: An examination of contingencies. Criminal Justice Studies, 17, 291-310.

Jensen, G. F. (1999). A critique of control balance theory: Digging into details. Theoretical Criminology, 3, 339-343.

Katz, J. (1988). Seductions of crime. New York: Basic Books.

Kim, M., a Hunter, J. (1993). Relationships among attitudes, behavioral intentions, and behavior: A meta-analysis of past research. Part 2. Communications Research, 20, 331-364.

Peters, L., a Murphy, A. (1993). Adolescent health survey: Report for the greater Vancouver region of British Columbia. Vancouver, BC: The McCreary Centre Society.

Piquero, A. R., a Hickman, M. (1999). An empirical test of Tittle’s control balance theory. Criminology, 37, 319-341.

Piquero, A. R., a Hickman, M. (2002). The rational choice implications of control balance theory. In A. R. Piquero & S. G. Tibbetts (Eds.) Rational choice and criminal behavior (pp. 85-107). New York: Routledge.

Piquero, A. R., a Hickman, M. (2003). Extending Tittle’s control balance theory to account for victimization. Criminal Justice and Behavior, 30, 282-301.

Piquero, A. R., Macintosh, R., a Hickman, M. (2001). Applying Rasch modeling to the validity of a control balance scale. Journal of Criminal Justice, 29, 493-505.

Shane, P. (1996). What about America’s homeless children? Thousand Oaks, CA: Sage.

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

Tittle, C. R. (1997). Thoughts stimulated by Braithwaite’s analysis of control balance theory. Theoretical Criminology, 1, 99- 110.

Tittle, C. R. (2001). Control balance. In R. Paternoster a R. Bachman (Eds.), Explaining criminals and crime (pp. 316-334). Los Angeles: Roxbury.

Tittle, C. R. (2004). Refining control balance theory. Theoretical Criminology, 8, 395-428.

Whitbeck, L. B., a Hoyt, D. R. (1999). Nowhere togrow. Hawthorne, NY: Walter de Gruyter.

Stephen W. Baron is an Associate Professor in the Department of Sociology at Queen’s University. His research interests focus on homeless street youth and crime. His work has recently appeared in Criminology, Journal of Research in Crime and Delinquency, and Journal of Criminal Justice.

David R. Forde is a Professor of Criminology and Criminal Justice and Director of the Mid-South Social Survey at the University of Memphis. His research interests include interpersonal victimization processes, crime prevention, and innovations in survey research methods. His recent publications have appeared in Violence & Victims, Journal of Interpersonal Violence and Child Abuse & Neglect. Correspondence to: Department of Sociology, Queen’s University, Kingston, Ontario, Canada K7L 3N6. E-mail: [email protected]

Comments

comments