Canadian Journal 1 to 2 Revenue canadienne de criminologie January/janvier 2000
On November 30th, 1994, the government of Ontario implemented the Ontario Tobacco Control Act (OTCA). The OTCA was legislated to address the issues of smoking in public places, related signage, and tobacco sales. The Act raised the legal age for selling tobacco from 18 to 19 years of age. The rationale for legislating a minimum age to purchase tobacco is based on research indicating that the onset of long term tobacco addiction occurs primarily during adolescence (Forster, Komro, and Wolfson 1996; Keay, Woodruff, Wildey, and Keeney 1993; Feighery, Altman, and Shaffer 1991; Altman, Rasenick-Douss, Foster, and Tye 1991). As few people begin smoking after the age of 20, tobacco prevention methods are designed to delay the initiation of tobacco consumption and reduce access to tobacco products for youth.
For merchants who do sell tobacco products illegally to underage youth, the OTCA set a graduated penalty structure. Fines range from a maximum fine of $2,000 for individuals who have committed a first offence, to a maximum of $25,000 for corporations convicted of three or more past offences. Such punishments for selling to underage minors in Ontario are among the most severe in North America.
Despite this new law and fine structure, tobacco products remain relatively easy for youth to access in Ontario, as many retailers remain willing to sell to minors. Research undertaken in Ontario in 1995 indicated that 27% of merchants were willing to sell cigarettes to minors (Abernathy 1996). At a national level, the Canadian Cancer Society (1995), for instance, measured the willingness of 499 retailers throughout Canada to sell cigarettes to minors. Overall, 60% of retailers tested were willing to. sell. Similarly, findings from the United States note that minors are able to purchase cigarettes at a success rate ranging from 46 to 76% (Altman, Foster, Rasenick-Douss, and Tye 1989; Altman, Rasenick-Douss, Foster, and Tye 1991; Cohen, Stanley, Martin, and Goldstein 1995; DiFranza, Norwood, Garner, and Tye 1987; Forster, Hourigan, and McGovern 1992; Jason, Ji, Anes, and Birkhead 1991; Roeseler, Caopra, and Quinn 1994; Wakefield, Carrangis, Wilson, and Reynolds 1992).
While it is well known that youth access to tobacco remains a problem, what is less clear is why so many merchants break the law and illegally sell tobacco products to minors. Data collected from 439 merchant compliance checks (events where underage youth attempt to purchase tobacco products) in Ontario are used to address this question.
Illegal tobacco sales to youth
The illegal sales of tobacco products to minors involves the interaction of two main players: the young person who wishes to purchase cigarettes and the merchant who may or may not be willing to sell. Why youth chose to smoke tobacco has been the subject of much psychological, physiological and sociological research (Adlaf, Smart, Walsh, and Ivis 1994; Akers 1996; Gottfredson and Hirschi 1990; Wahlgren, Hovell, Slymen, Conway, Hofstetter, and Jones 1997). It is the goal of this paper to depart from such an analysis and focus on the factors that influence the decisions of merchants who are confronted by youth who wish to purchase tobacco. Simply put, why do some merchants comply, while others break the law?
To a large extent, the existing literature on the topic has been produced by epidemiologists. This research has identified factors associated with illegal sales of cigarettes to minors, such as operation type, sex of youth, and enforcement patterns (Biglan, Henderson, Humphrey, Yasui, Whisman, Black, and James 1995: Cohen et al. 1995; DiFranza et al. 1987; Erickson, Woodruff, Wildey, and Keeney 1993; Forster et al. 1992; Rigotti, Stoto, Bierer, Rosen, and Schelling 1993). Most of these results, however, have been based on bi-variate analyses carried out in the United States, and have not considered a systematic socio-legal conceptual framework. The goal of this paper is to undertake a multivariate analysis of merchant sales to minors in Ontario, theoretically informed by a rational choice perspective.
Rational choice models have received increased attention in the fields of criminology, sociology, and political science, in an effort to understand the influences on legal violators' choices to offend (Akers 1990; Paternoster 1989). These conceptions, based on the "expected utility" principle in economic theory, assume that crime is calculated and deliberate -- that criminals are rational actors. Individuals are seen to make rational decisions based on the extent to which they expect the choice to maximize their profits or benefits and minimize costs or losses. According to this perspective, crime is understood to be influenced by variations in opportunity, environment, target, and risk of detection -- each of which change depending on situational context and type of offense. Such models have worked to bridge the gap between structural theories of crime with event based explanations, in an effort to understand why an individual at a particular moment makes the choice to partake in an illegal activity. Rational choice theorists argue that criminology has been preoccupied with explaining individual criminality while neglecting explanations of the criminal event (Cornish and Clarke 1986). We are attempting to merge explanations of crime which focus upon the impacts of external social and economic (background) factors over which individuals have little or no control (e.g., population density, the business cycle, geographic local) with situational context which places emphasis on the factors specific to the illegal/criminal event.
The rational choice model has been applied to a wide range of different offences including tax evasion (Varma and Doob 1998), traffic offending (Corbett and Simon 1992), corporate crime (Paternoster and Simpson 1996), drunk driving, larceny, and sexual assault (Nagin and Paternoster 1994). We contend that the analysis of illegal tobacco sales to minors can also be understood within a rational choice framework. Indeed, it is well recognized that punishments may deter certain crimes and certain offenders. Some years ago, William Chambliss (1975) made the distinction between crimes that are instrumental acts and those that are expressive. Examples of instrumental offences include burglary, embezzlement, tax evasion, auto-theft, and other illegal activities which are directed to some material end. Expressive acts would include murder, assaults, and sex offences, in which the behaviour is an end in itself. Chambliss contends that the influence of punishment is greater in instrumental crimes because they generally involve some planning and weighing of risks. By contrast, expressive crimes are often hasty and emotional acts. These criminals are unlikely to be concerned with the future consequences of their actions.
Chambliss also raises the point that a distinction can be made between individuals who have relatively high commitment to crime as a way of life and those who do not. The former would include those who engage in crime on a regular or possibly professional basis (such as organized criminals and prostitutes). For them, the likelihood of punishment is a constant feature of their life, something they have learned to live with. On the other hand, a tax evader, an occasional shoplifter, or a merchant who sells tobacco to underage minors, likely does not view his/her behaviour as criminal and receives little, if any, group support for these acts. Fear of punishment may well be a deterrent for such low-commitment individuals, particularly if they have reason to believe that the behaviour in which they are involved is actively being enforced (Vago 1994).
On the basis of this distinction -- instrumental and expressive acts, and high and low commitment offenders -- Chambliss suggests that punishment will have the greatest impact in situations that involve low-committed individuals engaging in instrumental acts.
From this typology, we contend that merchants selling tobacco to minors are low-committed offenders involved in instrumental illegal activities and are expected to refrain from such activity if the perceived costs of selling tobacco to minors (receiving a fine or losing their tobacco license) exceed the benefits (the revenue generated from illegal sales). This general approach seems more consistent with the principles of the rational choice model of criminality.
On a methodological note, rational choice theory has generally been tested using one of two research techniques: 1) self-report surveys (e.g., Corbett and Simon 1992; Paternoster 1989; Varma and Doob 1998) or; 2) instruments containing hypothetical scenarios administered within laboratory settings (Nagin and Paternoster 1994; Paternoster and Simpson 1996). What is lacking in this empirical research are tests of models using data taken from first hand observations of criminal events. Since the data we employ in this paper are based on the systematic recording of illegal events -- merchants selling to minors -- this paper offers a unique contribution to criminological research which tests the assumptions of a rational perspective.
It is well established in the tobacco control literature that the best way to determine the willingness of retailers to sell to minors is through the use of retailer compliance checks (DiFranza 1996). Compliance checks involve the use of underage youth to test retailers' willingness to sell them cigarettes.
Organized and administered at the health unit level(2), trained volunteer youth ranging in age between 13 and 18 were sent in pairs to a sample of randomly selected retailers across the province of Ontario, and according to a consistent, prepared script, asked to buy cigarettes. One minor would look for appropriate signage in the store, while the other youth attempted to purchase cigarettes. Informed consent was obtained from each volunteer, and if they were under the age of 16, the consent of their parents. All compliance checks took place during November and December of 1995.
Merchants selected for the study were drawn using a multistage random sample stratified on health unit and community size. To determine the communities from which the retailers would be selected, the proportion of the health unit's population residing in settlements of different sizes was calculated. Then, using the 1991 Canadian Census, each community within a participating health unit was placed in its respective population category. Once these lists were compiled, communities designated for compliance checks were selected using a random numbering process (For a more detailed discussion on the sample, see Abernathy 1996).
The youth were recruited and trained for the compliance checks, and care was taken to make sure that there would be little doubt that minors were involved in the checks. The youth after training were put in pairs and accompanied to pre-selected sample sites by an adult escort. During the compliance check, the youth asked to buy a package of cigarettes. If the retailer refused they left the site. If the retailer was willing to make a sale, the volunteer responded that he/she did not have enough money for the purchase, left the site and returned to the adult escort. Vendors were considered willing to make a sale if they either entered the sale on the cash register or offered tobacco to the youth and then asked for money. If the retailer asked for the age or identification of the youth, they truthfully gave their correct age or that they had no identification; if they were not asked they did not volunteer this information. Regardless of retailer willingness to sell, there was no actual purchase by the youth.
After the sales event, the team completed a prepared data form that identified the age and sex of the volunteer team, retailer type (i.e., convenience store, grocery store), presence of signage, and results of the purchase attempt.
The sample for the study was designed so that it would give a reliable estimate of the sales of tobacco to minors for the province of Ontario. It was calculated that, in order to make estimates for the year (1995) within +/- 5% within 95%, confidence limits required a sample of 475 retailers. Given the fact that not all health units in the province participated in the study, the number of compliance checks carried out totaled 439. Below, we discuss the measurement of the dependent variable, merchant compliance. We then work our way through a series of variables used in our analysis.
Data and variables
A dummy variable was used to measure the outcome of each sales event. If a merchant was prepared to sell cigarettes to a youth volunteer team, the event was recorded positively. Alternatively, if the youth were turned down in their purchase attempt, the event was recorded as negative. This technique has been used extensively in past research as a measure of merchant compliance (Abernathy 1996).
In keeping with the assumptions of a rational choice model of offending, three sets of variables were used in our analysis: enforcement, background, and event factors. The table in Appendix A provides a list of variables in the study, their operationalization, and selected descriptive statistics.
Appendix A Definitions and descriptive statistics of variables used in the analysis
Variable Description Merchant tobacco 1 = sale occurred sales compliance 0 = sale did not occur Time of the day Morning 1 = Yes, 0 = No (6:00 to 11:59 a.m.)(*) Afternoon 1 = Yes, 0 = No (12:00 to 5:59 p.m.) Evening 1 = Yes, 0 = No (6:00 to 11:59 p.m.) Age Mean age of youth volunteer pair Gender Male dyad 1 = Yes, 0 = No Female dyad 1 = Yes, 0 = No(*) Mixed dyad 1 = Yes, 0 = No Legal compliance 0 = Fewer than all 3 conditions were met index 1 = Signs were posted, ID was requested, and Age was asked Operation type Restaurant 1 = Yes, 0 = No Convenience 1 = Yes, 0 = No Grocery 1 = Yes, 0 = No Gas station 1 = Yes, 0 = No Other 1 = Yes, 0 = No(*) Rural or Urban 0 = population 5,000 and over 1 = population under 5,000 Tobacco producing 0 = Compliance check in non-tobacco region producing region 1 = Compliance check in tobacco producing region Charge rate Number of charges per 100,000 population Variable Mean Merchant tobacco .28 sales compliance Time of the day Morning .14 Afternoon .63 Evening .23 Age 15.35 Gender Male dyad .30 Female dyad .36 Mixed dyad .35 Legal compliance index .56 Operation type Restaurant .08 Convenience .56 Grocery .09 Gas station .18 Other .09 Rural or Urban .19 Tobacco producing region .15 Charge rate 1.33
(*) reference category
On the assumption that merchants weigh the costs and benefits of selling tobacco to minors during a sales event, an individual would be less inclined to break the law if there is active enforcement activity in their community. To capture the effect of enforcement, we examined the number of charges laid against tobacco merchants for illegally selling tobacco to minors, within a given health unit during 1995.(3) The charge rate is a standardized continuous variable measuring the number of charges that were laid on tobacco merchants, per 100,000 population. Each community where a sales event took place was assigned the charge rate of its respective health unit. This variable has been associated with merchant sales in past research (Roeseler et al. 1994; Jason et al. 1991).
Operation Type: Restaurants, gas stations, convenience stores, grocery stores and "other" locations (i.e., sports arenas and recreation centres) comprise the operation types selected for our analysis. In total, they represent all possible retail locations where tobacco is legally available in Ontario. Previous research has indicated that operation type has been associated with merchant compliance (Erickson et al. 1993; Forster et al. 1992; Wakefield et al. 1992).
Rural/Urban: This variable is formed on the basis of the population size of the community where each compliance check took place. Rural communities were defined as those communities whose population, according to the 1991 Census, was under 5,000 inhabitants. All communities with a population of 5,000 or greater were classified as urban. This variable is included based on the notion that surveillance in small communities is more difficult since tobacco inspectors are likely known to merchants. Given these circumstances, merchants are more willing to sell to underage youth in small towns because they assume the risk of being charged is negligible. On the other hand, the anonymity of inspectors in urban areas may create an environment where selling to underage minors may pose a risk, thus making it more difficult for underage youth to purchase tobacco.
Tobacco Producing Region: This variable measures the effect of merchants being located in communities where tobacco contributes to the economy (i.e., tobacco producing region). The main question is whether merchants are more likely to sell cigarettes to minors in communities in which tobacco plays a significant economic role. Research in the United States (Roeseler et al. 1994) and Canada (Marino 1995) has indicated that tobacco sales to minors varies by geographic region. The economic influence of the tobacco industry offers one possible explanation.
Legal Compliance Index: This variable is a composite measure taking into account the following three factors: 1) the posting of signage indicating the legal age to purchase cigarettes; 2) whether volunteer youth were asked for their personal identification during each compliance check; and 3) whether youth volunteers were asked for their age at each compliance check. All three of these conditions are required under the OTCA. We contend that merchants who comply with these requirements of the OTCA are not inclined to sell tobacco products to minors.
Time of Sales Event: The time of day for each compliance check was recorded in the study. For this analysis, the time of day is divided into three distinct periods: morning, afternoon, and evening. The times used to create these categories are contained in the Table in Appendix A. We argue that youth who attempt to make a purchase during school hours are more suspect, and thus merchants are less likely to provide them with tobacco. On the other hand, youth purchase attempts that take place outside of school hours (i.e., during the evening) are more successful.
Age of Youth Volunteers: This variable measures the mean age of the youth volunteer teams partaking in the compliance checks. Two assumptions justify the inclusion of this variable in our analysis. On the one hand, merchants are likely to consider it morally/personally offensive to supply a 12 year old, as opposed to a 18 year old with cigarettes. On the other hand, and not unrelated to the last point, the perceived risk of selling tobacco to those who appear to be of legal age (19) is less than the risk of selling to youth who obviously appear to be under the legal age to purchase cigarettes.
Gender Composition of Youth Volunteer Teams: The gender composition of each youth volunteer team was established. Three dyads exist: 1) two females; 2) two males; and 3) one male and one female. Past research indicates that underage females have more success purchasing tobacco compared to their under age male counterparts simply because adolescent females look older, on average, than adolescent males (Forster et al. 1992; DiFranza et al. 1987; Cohen et al. 1995).
Data analysis is performed using a logistic regression procedure. This procedure estimates the coefficients of a probabilistic model, involving a set of independent variables, that best predicts the value of a dichotomous dependent variable (Hardy 1993; Wheeler, Weisburd, and Bode 1982).
Table 1 presents the results of the logistic regression analysis of illegal merchant sales of tobacco to minors. Results are reported using probability estimates for the statistically significant variables in our models.(4) We examine the results separately for each of the three models. Is the illegal sale of tobacco to youth related to enforcement? When considering the independent effect of charge rate on illegal sales, this variable is significant and shows a substantial effect on sales: the lower the charge rate the greater the likelihood of illegal sales. More specifically, when using probability estimates, when the charge rate in a health unit is zero the likelihood of illegal sales was 33.5 percent. On the other hand, illegal sales drop considerably as charge rates increase. With a charge rate of 10, the likelihood of illegal sales is one-sixth as high, 5.5 percent.
Logistic regression models for enforcement, background and event factors
Model one Logistic Anti log coefficient(1) odds (b)(2) Enforcement: Charge rate(3) -.221(**) 0.80 Background factors: Rural/urban Gas station Other Grocery stores Convenience stores Tobacco producing region Event factors: Afternoon Evening Age of children(4) Male/Male Female/Male Legal compliance Nagelkerke R2 R2 = .024 N = 439 Model two Model three Logistic Anti log Logistic Anti log coefficient odds (b) coefficient odds (b) Enforcement: Charge rate(3) -.219(*) 0.80 -0.05 0.95 Background factors: Rural/urban -.010 0.98 -0.82 0.44 Gas station .850 2.34 0.88 2.40 Other .994 2.70 1.35 3.88 Grocery stores .301 1.35 0.55 1.74 Convenience stores .462 1.58 0.15 1.17 Tobacco producing region -.256 0.77 0.16 1.12 Event factors: Afternoon 0.94 2.54 Evening 1.10(*) 3.02 Age of children(4) 0.93(**) 2.52 Male/Male -1.32(**) 0.25 Female/Male 0.90(*) 2.30 Legal compliance -4.07(**) 0.02 Nagelkerke R2 R2 = .046 R2 = .58 N = 437 N = 399
(*) Indicates significance (p = 0.05)
(**) Indicates significance (p = 0.01)
(1) The odds ratio for each variable estimates the likelihood that an event occurs: odds above "1" indicate an increased likelihood, and odds below "1" indicate a decreased likelihood of an event taking place.
(2) Charge rate and tobacco region are measured at an aggregate level. Though the remaining variables in our analysis are at the individual level, we were unable to obtain relevant data on charge rate and tobacco region at the individual (merchant) level.
(3) We also examined the possibility that the effect of age varies across gender dyads. We found no significant interaction effect between gender dyads and age of youth.
(4) Probability estimates measure the likelihood of merchant sales, in percentages, when scores on all other variables are held at their mean.
When our selected background factors -- type of business operation, tobacco producing region, and rural/urban location -- are introduced in Model 2 they do not affect illegal sales. At the same time, charge rate maintains a significant negative relationship to sales. In fact, as Table 1 points out, the strength of the effect of charge rate on sales is only marginally reduced (from 0.221 to 0.219) after the introduction of the background variables. Thus, the three background factors did not have a significant influence on illegal sales. After controlling for these variables, the original relationship remained statistically significant.
The strongest influence on illegal merchant tobacco sales results from the introduction of four event specific factors -- time of the day, age of the youth team, gender of the youth team, and legal compliance behaviour of the merchants. Model 3 indicates that the time of the day in which the sales event takes place significantly alters the likelihood of offending. If the purchase attempt occurs in the morning, the likelihood of sales is lowest (6 percent). Minors' success improves throughout the day, with afternoon sales greater than those attempted in the morning (18 percent). Finally, when youth attempt to purchase cigarettes after 6:00 p.m., their likelihood of success significantly increases to 21 percent.
Gender is also significantly related to illegal sales. Consistent with other research on illegal merchant tobacco sales reported earlier, the male dyad is less likely to be sold cigarettes with a probability of 6 percent. Female dyads, on the other hand, are more successful in their purchase attempts (14 percent). The one surprising finding involves the mixed dyad. We anticipated that a dyad consisting of both a male and female would be situated between the male and female dyads. Instead, the likelihood of sales is the highest at 22 percent, when the dyad is of mixed gender.
Age of the youth is also significantly related to the probability that merchants will sell illegally. The age of the youth involved in the study ranged from 13 to 18 years old. The age of the youth strongly influences the probability of sales (p = 0.001). As the age of youth volunteers increases, their purchase attempts are more successful. For example, the probability of sales for youth aged 13 is 2 percent, which rises to 10.5 percent for 15 year olds and to 65% for 18 year olds.
The strongest predictor in model 3 is legal compliance. When tobacco merchants comply with other tobacco related laws (posting signs, asking for identification and age of youth), the likelihood of illegal sales is relatively low (2.5 percent). Conversely, when compliance with any one of these conditions is not met, the probability of merchants offending rises (61 percent).
Finally, how has the introduction of event factors influenced the impact of charge rate on merchant offending? Once event factors are introduced into the model, the relationship between illegal merchant sales and charge rate became non-significant (p = .66). In other words, the relationship between sales and charge rate is in fact suppressed, such that its impact has been removed once event specific variables were considered.
Discussion and conclusion
The findings in this article lend support for a rational choice model of illegal sales of tobacco products to underage youth. This rational choice explanation, however, places a strong emphasis on the situational factors involved with illegal events. Decision makers' assessment of the costs and benefits of such actions during the offense affect the choices involved in selling cigarettes to minors. Though the likelihood of sales at the bivariate level, as well as when controlling for background factors, are related to local enforcement activity, this deterrent effect weakens once event factors are introduced into the model. These findings support the earlier discussion that merchants selling tobacco to minors are low-committed offenders involved in instrumental deviant acts. It is therefore unlikely that much planning or weighing of risks is involved a priori on behalf of merchants. Rather, an important cognitive risk assessment occurs at the point when a merchant is confronted with the opportunity to sell tobacco to a young person. This explains why the age (older youth) and gender (females -- who generally mature earlier than males) composition of the youth teams were so important for explaining the likelihood of an illegal sale. These findings support the assumption that, if they have reason to believe that a youth "appears" to be of legal age to purchase tobacco, the merchants are more likely to supply the product. This finding has been well documented in several American studies (Roeseler et al. 1994; Forster et al. 1992; DiFranza et al. 1987; Cohen et al. 1995).
The important role played by our situational factors can be related to recent research on youth crime and homelessness. Hagan and McCarthy (1997) argue that crime committed by street youth is more related to situational pressures, such as hunger and homelessness, than background and developmental factors -- experiences that take place in the family, at school, and among peers. Similarly, this research supports an emphasis on the study of criminogenic situations, rather than the role of predispositions and propensities in the study of legal wrong doing.
The finding that successful purchase attempts are associated with evening activity can be explained on the basis of enforcement patterns. Since tobacco inspectors assigned to health units across Ontario routinely work during regular daytime business hours (i.e., 9:00-5:00), merchants selling to minors are at a greater risk of being changed during the day than in the evening.(5) While this research design was unable to collect information on the characteristics of the sales staff (e.g., age and gender), it may well be the case that younger staff are more likely to work during evenings and could be more sympathetic to underage smokers. This would be an ideal focus for future research.
Merchants who follow other conditions imposed by the OTCA (required signage, asking for age and ID) are found to be the least likely to sell to underage youth. This may imply that vigilant attention by inspectors ensuring the proper posting of signs and checking for ID would decrease illegal sales of tobacco. In addition to enforcement, it may not be unreasonable to suggest that legal conformity is related to an individual's own personal moral beliefs about breaking the law or feelings of shame or guilt. Indeed, these elements have proven to be useful in other research testing rational choice models (Paternoster and Simpson 1996). It would be a task for future research on tobacco control to attain more information on the backgrounds and attitudes of tobacco vendors.(6)
Finally, these findings are not without policy implications. First, to avoid the temptation of merchants who are willing to sell tobacco to youth who may `look to be 19', legislation could be invoked whereby any person who appears to be under the age of 25, for example, would be required to produce personal identification before being permitted to purchase tobacco. Such a strategy is currently in place to regulate the sale of tobacco to `minors' in the United States. The Food and Drug Administration has set a policy which restricts tobacco sales to those under the age of 18, but requires identification for those who appear to be under the age of 27. Second, if enforcement practices were to be extended into the evening hours, we would hypothesize that it would have deterrent effect upon merchants' behaviour. Or alternatively, given the shrinking budgets of Ontario's health units, it may be wise to divert funds from day-time to night-time inspection.
In conclusion, given our focus upon `real life' illegal events, and the powerful explanatory role of the event factors in our analysis, we hope that this paper has made a useful contribution to research evaluating the rational choice perspective, and efforts to control youth access to tobacco.
(1.) Address correspondence to: Tom Abernathy, Director, Central West Health Planning Information Network, 10 George St., #301B, Hamilton, ON L8P 1C8, (905) 570-9952, ext. 239, firstname.lastname@example.org.
(2.) As of 1996, the province of Ontario was divided into 42 health units, determined by geography and administrative efficiency. Health units are responsible for monitoring public health policy and research for the provincial government.
(3.) Charge data are only collected at the health unit level by the Ontario Ministry of Health. We were unable to obtain the names of the specific merchants in which charges were laid. As such, charge rate is an aggregate level variable and does not account for prior fines assigned to a specific merchant.
(4.) A statistic measuring the amount of variation in the dependent variable that is due to changes in the independent variable.
(5.) While police in Ontario also have the power to charge merchants under the OTCA, to date the vast majority of charges levied against merchants in Ontario have been laid by health unit inspectors (Personal communication with Tom Abernathy).
(6.) We did attempt to gather information on the backgrounds and attitudes of merchants who were involved in our sales events. One week after each sales event, merchants were contacted and asked to participate in a telephone survey. For merchants who were willing and able to respond to our survey, they were asked attitudinal and background questions related to tobacco sales and control. Unfortunately, due to a low response rate and missing data, this information not able to be included in our analysis.
1996 A Study to Measure the Impact of the Tobacco Control Act. Toronto: Ontario Ministry of Health.
Adlaf, E., R. Smart, G. Walsh, and F. Ivis
1994 Is the association between drug use and delinquency weakening? Addiction 89: 1675-1681.
1990 Rational choice, deterrence, and social learning theory in criminology: The path not taken. The Journal of Criminal Law and Criminology 81(3): 653-76.
1996 A longitudinal test of social learning theory. Journal of Drug Issues 26(2): 317-343.
Altman, D., V. Foster, L. Rasenick-Douss, and J. Tye
1989 Reducing the illegal sale of cigarettes to minors. Journal of the American Medical Association 261(1): 80-83.
Altman, D., L, Rasenick-Douss, V. Foster, and J. Tye
1991 Sustained effects of an education program to reduce sales of cigarettes to minors. American Journal of Public Health 81(7): 891-893.
Biglan, A., J. Henderson, D. Humphrey, M. Yasui, R. Whisman, C. Black, and L. James
1995 Mobilizing positive reinforcement to reduce youth access to tobacco. Tobacco Control (4): 42-48.
Canadian Cancer Society
1995 Sixty percent of retailers sell cigarettes to kids. New release, September 25.
1975 Types of deviance and the effectiveness of legal sanctions. In William Chambliss (ed.), Criminal Law in Action. Santa Barbara, California: Hamilton Publishing Company.
Cohen, J., L. Stanley, J. Martin, and A. Goldstein
1995 Illegal sales of cigarettes to minors in North Carolina. North Carolina Medical Journal 56(1): 59-63.
Corbett, C. and F. Simon
1992 Decisions to break or adhere to the rules of the road, viewed from a rational choice perspective. British Journal of Criminology 30(4): 537-49.
Cornish, D. and R. Clarke (eds.)
1986 The Reasoning Criminal: Rational Choice Perspectives on Offending. New York: Springer-Verlag.
DiFranza, J., B. Norwood, D. Garner, and J. Tye
1987 Legislative efforts to protect children from tobacco. Journal of the American Medical Association 257(24): 3387-3389.
1996 Active enforcement of minors' access laws: A moral and ethical imperative. Tobacco Control 4:361.
Erickson, A., S. Woodruff, M. Wildey, and E. Keeney
1993 A baseline assessment of cigarette sales to minors in San Diego, California. Journal of Community Health 18(4): 213-24.
Feighery, E., D. Altman, and G. Shaffer
1991 The effects of combining education and enforcement to reduce tobacco sales to minors. A study of four Northern California communities. Journal of the American Medical Association. 266(22): 3168-3171.
Forster, J., M. Hourigan, and P. McGovern
1992 Accessibility of cigarettes to underage youth in three communities. Preventive Medicine 21: 320-328.
Forster, J., K. Komro, and M. Wolfson
1996 Survey of city ordinances and local enforcement regarding commercial availability of tobacco to minors in Minnesota, United States. Tobacco Control 5: 46-51.
Gottfredson, M. and T. Hirschi
1990 A General Theory of Crime. Palo Alto, CA: Stanford University Press
Hagan, J. and B. McCarthy
1997 Mean Streets: Youth Crime and Homelessness. Cambridge University Press: New York.
1993 Regression with Dummy Variables. Sage University paper series. Quantitative applications in the social sciences (93). Newbury Park, California: Sage.
Jason, L., P. Ji, M. Anes, and S. Birkhead
1991 Active enforcement of cigarette control laws in the prevention of cigarette sales to minors. Journal of the American Medial Association 266(22): 3159-3161.
Keay, K., S. Woodruff, M. Wildey, and E. Keeney
1993 Effect of a retailer intervention on cigarette sales to minors in San Diego County, California. Tobacco Control 3: 145-51.
1995 Measure of Retailer Compliance with respect to Tobacco sales-to-Minors-Legislation and Restrictions on Tobacco Advertising. Prepared for Health Canada - Office of Tobacco Control. Toronto: Ontario Ministry of Health. Ontario Health Survey.
Nagin, D. and R. Paternoster
1994 Personal capital and social control: The deterrence implications of a theory of individual differences in criminal offending. Criminology 32(4): 581-602.
1989 Absolute and restrictive deterrence in a panel of youth: Explaining the onset, persistence/desistence, and frequency of delinquency offending. Social Problems 36(3): 291-309.
Paternoster, R. and S. Simpson
1996 Sanction threats and appeals to morality: Testing a rational choice model of corporate crime. Law and Society Review 30(3): 540-580.
Rigotti, N., M. Stoto, M. Bierer, A. Rosen, and T. Schelling
1993 Retail stores' compliance with a city no-smoking law. American Journal of Public Health 83(2): 227-32.
Roeseler, A., A. Caopra, and V. Quinn
1994 A special campaign to generate support for state legislation to decrease youth access to tobacco: A case study. Tobacco Control Section, California Department of Health Services. American Public Health Association Meeting, Washington, D.C., November 2.
The Ontario Tobacco Control Act (Bill 119) (OTCA)
1993 An Act to prevent the Provision of Tobacco to Young Persons and to Regulate its Sale and Use by Others. Legislative Assembly of Ontario.
1994 Law and Society (4th Ed.) Englewood Cliffs, New Jersey: Prentice Hall.
Varma, K. and A. Doob
1998 Deterring economic crimes: The case of tax evasion. Canadian Journal of Criminology 40(2): 165-84.
Wakefield, M., J. Carrangis, D. Wilson, and C. Reynolds
1992 Illegal cigarette sales to children in South Australia. Tobacco Control 1: 114-117.
Wahlgren, D., M. Hovell, D. Slymen, T. Conway, C. Hofstetter, and J. Jones
1997 Predictors of tobacco use initiation in adolescence: A two-year prospective study and theoretical discussion. Tobacco Control 6: 95-103
Wheeler, S., D. Weisburd, and N. Bode
1982 Sentencing the white-collar offender: Rhetoric and reality. American Sociological Review 47:641-59.
William O'Grady Department of Sociology and Anthropology University of Guelph Guelph, Ontario Mark Asbridge Department of Sociology University of Toronto Toronto, Ontario Tom Abernathy Central West Health Planning Information Network and Ontario Tobacco Research Unit Hamilton, Ontario…