Abstract
Objectives. We assessed the effect of internal possession (IP) laws, which allow law enforcement to charge underage drinkers with alcohol possession if they have ingested alcohol, on underage drinking behaviors.
Methods. We examined Youth Risk Behavior Survey (YRBS) data from 12 states with IP laws and with YRBS data before and after each law’s implementation. We used logistic regression models with fixed effects for state to assess the effects of IP laws on drinking and binge drinking among high school students.
Results. Implementation of IP laws is associated with reductions in the odds of past-month drinking. This reduction was bigger among male than among female adolescents (27% vs 15%) and only significant among younger students aged 14 and 15 years (15% and 11%, respectively). Male adolescents also reported a significant reduction (24%) in the odds of past-month binge drinking under IP laws.
Conclusions. These findings suggest that IP laws are effective in reducing underage drinking, particularly among younger adolescents.
The problem of youth drinking is hardly new. The US government first took steps to address this public health issue in 1984, with the passage of a minimum legal drinking age (MLDA)1 that required states to prohibit the purchase and possession of alcohol by people younger than 21 years or risk losing vital highway funds. By 1988, all 50 states had complied.2 In the decades after the enactment of the MLDA, states and localities have implemented numerous additional policies in an attempt to reduce underage drinking. Keg registration laws, server training requirements, compliance checks performed at retail outlets, and distinctive licenses for drivers younger than 21 years are all examples of policies that seek to restrict alcohol’s availability to youths, but their breadth and enforcement vary widely from state to state. Although all states prohibit the possession of alcohol by a minor, for example, only 35 states have laws that prohibit consumption by a minor.3 Fell et al.4 examined the possession prohibition, as well as 5 other underage drinking laws. They found that a decrease in the ratio of underage drinking drivers to nondrinking drivers in fatal crashes was associated with the MLDA provisions prohibiting purchase and possession of alcohol by minors; zero tolerance laws, which make it illegal for minors to operate a vehicle with any amount of alcohol in their system; and use-and-lose laws, which allow for the suspension of driving privileges for minors guilty of alcohol violations.
Closely linked with possession and consumption laws, internal possession (IP) laws prohibit minors from possessing alcohol within their bodies. This policy emerged in response to the difficulty law enforcement encountered in citing youths for violations at underage drinking parties. As 1 newspaper article described,
In the old days, they say, the teenagers at a party would drop their drinks and run when officers arrived. That would leave the police with few of the particulars—who drank what, and when—necessary to build a legal case.5
Without witnessing an underage person possessing or consuming alcohol, law enforcement was left with little evidence with which to charge an underage drinker. IP laws address this issue by allowing for underage drinkers to be cited if law enforcement determines that they have been drinking either by observing outward signs of intoxication or through the use of an objective measure such as a blood, breath, or urine test. Eight states require objective proof: Colorado, Kansas, Michigan, Missouri, New Hampshire, North Carolina, South Carolina, and Utah. In this article, we refer to those states with laws requiring only outward signs of intoxication as subjective proof states: Arizona, Idaho, Nebraska, North Dakota, Vermont, and Delaware.3
As part of its effort to provide detailed information on a wide variety of alcohol-related policies in the United States at both the state and the federal levels, the Alcohol Policy Information System (APIS), funded by the National Institute on Alcohol Abuse and Alcoholism, has tracked IP laws along with 34 other relevant policies. However, research on the effectiveness of IP laws on reducing underage drinking has been limited. Fell et al.6 assessed the relationship of 16 key underage drinking laws to reductions in underage drinking drivers involved in fatal traffic crashes. In that study, IP laws were included as 1 component of possession laws but were not examined separately. Of the 16 laws examined, only the prohibition of using false identification was associated with the reduction of underage drinking drivers in fatal crashes. A second analysis focusing on purchase and possession prohibitions found that the presence of these 2 laws was associated with a decrease in the ratio of underage drinking drivers to underage nondrinking drivers in fatal crashes.
We sought to fill this gap in the literature by evaluating IP laws using data from state-based surveys of high school students. Our objective was to assess the effect of IP laws on underage drinking behaviors.
METHODS
Our data are drawn from the Youth Risk Behavior Surveillance System, which was developed in 1990 to monitor priority health risk behaviors among US high school students. The system includes national, state, territorial, tribal, and local school-based surveys of representative samples of high school students. The national survey is conducted by the Centers for Disease Control and Prevention, and the rest are conducted by departments of health or education in each jurisdiction. These surveys have been conducted every 2 years since 1991. For this study, the data were taken from states’ Youth Risk Behavior Survey (YRBS). Each state conducts the survey in high schools in every odd-numbered year between February and May, using a 2-stage cluster design to produce representative samples of students in grades 9 through 12 in their jurisdiction. Data are collected through student self-administered questionnaires. Surveys that have a scientifically selected sample, appropriate documentation, and an overall response rate of at least 60% are weighted to represent the high school student populations of the state. Although all states are eligible to apply for YRBS funding, in 1991 only 9 states participated. By 2009, only 1 state (Minnesota) had never conducted the survey, and 4 others (California, Oregon, Virginia, and Washington) had no weighted data. Not all states conduct the survey in every biennium. In sum, 45 states and the District of Columbia have weighted data for at least 1 year. For the study objective, we included in the analysis only states with IP laws for which weighted data were available for the years before and after each law’s enactment.
We used APIS to obtain the dates of enactment for all laws after 1998. For laws enacted before 1998, we searched the regulatory database of each state to identify the date of enactment. APIS considers a state to have an IP law if it has a statute or regulation prohibiting a minor from having alcohol in his or her system as determined by a blood, breath, or urine test. The database further identifies 6 states with laws that punish minors for displaying “indicators of consumption” or for “exhibiting the effects” of having consumed alcohol without the requirement of an objective test, although APIS does not categorize these as IP laws. After conducting analyses on the 2 types of IP laws separately and finding similar patterns of results, we included both types of laws in a single category in this study. Although Colorado and Delaware have IP laws in effect, we excluded them from the analysis because preimplementation data were not available. Therefore, our analysis included data for 12 states.
Measures
We measured 2 current drinking behaviors by 2 alcohol-related questions asked by all states:
Any drinking in the past month: During the past 30 days, on how many days did you have at least 1 drink of alcohol?
Binge drinking in the past month: During the past 30 days, on how many days did you have 5 or more drinks of alcohol in a row—that is, within a couple of hours?
We used these measures as outcome variables and dichotomized them (yes or no) so that each respondent was categorized as either having engaged in the behavior at least once in the past month or not. Measures of IP laws were dichotomously coded so that each state either had the law in effect in a given year (coded 1) or did not (coded 0).
Keg registration laws, social host liability laws, and laws prohibiting the consumption of alcohol by minors were enacted by several states in the 1990s and 2000s. Keg registration laws are intended to reduce underage drinking by limiting access to alcohol among minors because parties with kegs provide minors with easy access to cheap or free beer.7 However, research on these laws has generated mixed results. Ringwalt and Paschall8 found a moderate but not statistically significant reduction in adolescent binge drinking associated with keg registration laws. Research by Cook9 indicated that states with keg registration laws have higher rates of underage drinking. A study by Fell et al.4 showed a reduction in beer consumption but an increase in underage drinking-and-driving fatal crashes. Other studies of traffic deaths have produced similarly mixed results.10 Research on social host liability laws and consumption prohibitions (separate from other MLDA provisions) has been minimal. Nonetheless, we included these laws in our models to control for any potential impact they may have had on the outcomes.
We coded states as having a keg registration law in a given year if they had policies in effect that prohibited possessing an unregistered or unlabeled keg or required the retailer to collect the purchaser’s name and address or identification number from a government ID. Because Utah prohibits the sale of kegs entirely, we coded Utah as having this policy for all years during the study period.
Other control variables included demographic characteristics of respondents, namely age (≤ 14, 15, 16, 17, and ≥ 18 years), gender, and race/ethnicity. Respondents younger than 14 years were combined with 14-year-old respondents to form 1 age group because there were few respondents in the younger age group. Because of small numbers of respondents in some minority groups, we coded race/ethnicity in 4 categories: White non-Hispanic, Black non-Hispanic, Hispanic, and other (including Asian, American Indian/Alaska Native, Native Hawaiian/Pacific Islander, or other). In addition, we created dummy variables to represent each of the respondent states. Year of survey was used as a continuous variable to represent time trends.
Statistical Analysis
Data from all 12 IP law states and all survey years were pooled into a single sample. We used a logistic regression model to assess whether state IP laws had a significant effect on each of the drinking-related outcomes while controlling for demographic characteristics of students, other underage drinking law changes, state, and survey year. We included dummy variables of states as covariates in an attempt to capture the fixed effect of unmeasured state characteristics and thus to isolate the within-state prelaw and postlaw comparisons. Year of survey was intended to control for general time trends in the outcomes. To examine the potential differential effects of the policy across subpopulation groups, we tested interactions between the IP law and age, gender, and race/ethnicity in each model. Because the interaction with race/ethnicity was not statistically significant, we did not retain it in the final models. For binge drinking, we ran 2 models: 1 among all respondents and 1 among respondents who reported any drinking in the past month. We used the survey logistic regression procedure in SAS version 9.2 (SAS Institute, Inc., Cary, NC) to account for sampling weights and sampling design effects.
We excluded a small proportion of respondents (1.7%) missing 1 or more of the demographic variables from the analyses, resulting in a total of 219 171 students in the analytic sample. Additional cases had missing values on the outcome variables. To avoid losing too many cases, we excluded only respondents who were missing a specific outcome variable from that particular model. We were able to recode some missing responses on the basis of responses on the other outcome. For example, respondents who reported that they had not had a drink in the previous 30 days were recoded as “no” on the binge drinking outcome if they were missing on that variable. Conversely, respondents who reported that they had engaged in binge drinking were recoded as “yes” on the any drinking outcome if they were missing on the variable. Once this recoding process was complete, 8906 responses (4.1%) remained as missing values and were excluded from the any drinking model and 5150 (2.4%) were excluded from the binge drinking model.
For each underage drinking policy, we included a lag time of 1 year from enactment to allow for any effects of the policy to emerge. Because the YRBS is only conducted in odd-numbered years, we did not code a state that enacted an IP law in an odd-numbered year as having the law until the following YRBS data year. If the state enacted the policy in an even-numbered year, we used the exact date to determine whether the YRBS was conducted at least 1 full year after implementation. March 1 was assumed to be the date of administration of the YRBS in each odd-numbered year, so a state that enacted a policy in July 2000 would not be coded as having that policy in effect until the 2003 YRBS data year. The states included in the study and the effective dates of the underage drinking laws of interest are presented in Table 1. When no exact enactment date was available for policies enacted in even-numbered years, we assumed that the policy was enacted after the March 1 cutoff date, ensuring at least 1 year between enactment and the YRBS survey.
TABLE 1—
Enactment Dates of Underage Drinking Policies and Available Youth Risk Behavior Survey Data by State: 1991–2009
| State | Internal Possession | Consumption | Keg Registration | Social Host | YRBS Data Years |
| Arizona | May 2002 | 1998a | … | 1998a | 2003–2009 |
| Idaho | July 2000 | 1984b | July 1981 | … | 1991–1993, 2001–2009 |
| Kansas | July 2006 | 1985b | July 2002 | July 2004 | 2005–2009 |
| Michigan | April 1998 | April 1998 | … | March 1994 | 1997–2009 |
| Missouri | August 2005 | … | July 2004 | August 2005 | 1995–2009 |
| Nebraska | April 2001 | September 2001 | June 1993 | … | 1991–1993, 2003–2005 |
| New Hampshire | January 2003 | … | February 2001 | April 2004 | 1993–1995, 2005–2009 |
| North Carolina | December 2006 | December 2006 | December 2006 | … | 1993–1995, 2001–2009 |
| North Dakota | August 1999 | 1987b | July 1983 | … | 1995, 1999–2009 |
| South Carolina | July 2007 | July 2007 | January 2008 | July 1994 | 1991–1999, 2005–2009 |
| Utah | May 2004 | 1935b | NA | May 2009 | 1991–2009 |
| Vermont | July 2000 | July 2000 | July 1992 | … | 1999–2009 |
Note. YRBS = Youth Risk Behavior Survey; NA = not applicable. Ellipses indicate that the state does not currently have the specified law.
The Alcohol Policy Information System indicated that the law was in effect as of January 1 of the specified year. We were unable to determine an exact date of enactment.
We were unable to determine the specific month of enactment.
RESULTS
Table 2 provides demographic characteristics of the YRBS respondents in IP states in the periods before and after the law was implemented. Because grade and age are highly correlated, we included only age as a control in logistic regression models.
TABLE 2—
Distribution of Respondents Before and After the Implementation of Internal Possession Laws: Youth Risk Behavior Survey, 1991–2009
| Variable | Before IP Law, No. (%) | After IP Law, No. (%) | Total No. (%) |
| Total | 127 093 (100) | 92 078 (100) | 219 171 (100) |
| Gender | |||
| Male | 62 556 (49) | 45 896 (50) | 108 452 (49) |
| Female | 64 537 (51) | 46 182 (50) | 110 719 (51) |
| Race/ethnicity | |||
| White, non-Hispanic | 93 509 (74) | 71 534 (78) | 165 043 (75) |
| Black, non-Hispanic | 20 588 (16) | 5308 (5) | 25 896 (12) |
| Hispanic | 4678 (4) | 6999 (8) | 11 677 (5) |
| Other | 8318 (7) | 8237 (9) | 16 555 (8) |
| Age, y | |||
| ≤ 14 | 14 592 (11) | 11 059 (12) | 25 651 (12) |
| 15 | 33 756 (27) | 24 613 (27) | 58 369 (27) |
| 16 | 34 490 (27) | 24 570 (27) | 59 060 (27) |
| 17 | 28 649 (23) | 20 700 (22) | 49 349 (23) |
| ≥ 18 | 15 606 (12) | 11 136 (12) | 26 742 (12) |
| State | |||
| Arizona | 1969 (2) | 5107 (6) | 7076 (3) |
| Idaho | 9898 (8) | 6664 (7) | 16 562 (8) |
| Kansas | 3339 (3) | 2002 (2) | 5341 (2) |
| Michigan | 6504 (5) | 16 986 (18) | 23 490 (11) |
| Missouri | 12 722 (10) | 3126 (3) | 15 848 (7) |
| Nebraska | 5607 (4) | 6655 (7) | 12 262 (6) |
| New Hampshire | 6053 (5) | 4309 (5) | 10 362 (5) |
| North Carolina | 16 532 (13) | 5587 (6) | 22 119 (10) |
| North Dakota | 3305 (3) | 8486 (9) | 11 791 (5) |
| South Carolina | 28 105 (22) | 1077 (1) | 29 182 (13) |
| Utah | 19 099 (15) | 3481 (4) | 22 580 (10) |
| Vermont | 13 960 (11) | 28 598 (31) | 42 558 (19) |
| Past-mo drinking behaviors | |||
| Any drinking | |||
| Yes | 52 804 (43) | 36 550 (41) | 89 354 (43) |
| No | 69 334 (57) | 51 577 (59) | 120 911 (58) |
| Binge drinking | |||
| Yes | 32 355 (26) | 23 366 (26) | 55 721 (26) |
| No | 91 279 (74) | 67 021 (74) | 158 300 (74) |
| Binge drinking among current drinkers | |||
| Yes | 32 355 (63) | 23 366 (65) | 55 721 (64) |
| No | 18 894 (37) | 12 827 (35) | 31 721 (36) |
Note. IP = internal possession. Percentages may not add to 100% because of rounding.
Table 3 presents the odds ratios from 6 logistic regression models of 2 drinking behaviors: any drinking in the past month and binge drinking in the past month. For each outcome, we present results with and without interaction terms included in the model.
TABLE 3—
Odds Ratios for Any Drinking and Binge Drinking in the Past Month: Youth Risk Behavior Survey, 1991–2009
| Any Drinking (n = 210 265) |
Binge Drinking (n = 214 021) |
Binge Drinking Among Current Drinkers (n = 87 442) |
||||
| Independent Variables | Model 1, OR (95% CI) | Model 2, OR (95% CI) | Model 3, OR (95% CI) | Model 4, OR (95% CI) | Model 5, OR (95% CI) | Model 6, OR (95% CI) |
| Age, y | ||||||
| ≤ 14 (Ref) | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| 15 | 1.20*** (1.13, 1.28) | 1.17*** (1.09, 1.26) | 1.26*** (1.18, 1.35) | 1.20*** (1.10, 1.31) | 1.16*** (1.08, 1.25) | 1.11 (0.99, 1.25) |
| 16 | 1.60*** (1.50, 1.71) | 1.52*** (1.40, 1.65) | 1.75*** (1.63, 1.88) | 1.66*** (1.53, 1.81) | 1.41*** (1.31, 1.53) | 1.40*** (1.26, 1.56) |
| 17 | 1.92*** (1.80, 2.05) | 1.78*** (1.64, 1.93) | 2.31*** (2.16, 2.47) | 2.14*** (1.96, 2.34) | 1.81*** (1.66, 1.98) | 1.77*** (1.57, 2.00) |
| ≥ 18 | 2.33*** (2.14, 2.52) | 2.08*** (1.89, 2.28) | 2.76*** (2.55, 2.98) | 2.45*** (2.23, 2.70) | 1.95*** (1.79, 2.12) | 1.84*** (1.63, 2.07) |
| Race/ethnicity | ||||||
| White, non-Hispanic (Ref) | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Black, non-Hispanic | 0.61*** (0.58, 0.64) | 0.60*** (0.58, 0.64) | 0.36*** (0.34, 0.39) | 0.37*** (0.34, 0.39) | 0.37*** (0.34, 0.41) | 0.37*** (0.34, 0.41) |
| Hispanic | 1.21*** (1.13, 1.29) | 1.21*** (1.13, 1.29) | 1.13** (1.06, 1.21) | 1.13** (1.06, 1.21) | 1.03 (0.92, 1.16) | 1.03 (0.92, 1.16) |
| Other | 0.95 (0.89, 1.02) | 0.96 (0.89, 1.03) | 0.95 (0.88, 1.03) | 0.95 (0.88, 1.04) | 1.05 (0.96, 1.15) | 1.05 (0.96, 1.15) |
| Gender | ||||||
| Female (Ref) | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Male | 1.06* (1.03, 1.10) | 1.13*** (1.08, 1.18) | 1.26*** (1.22, 1.30) | 1.34*** (1.28, 1.41) | 1.48*** (1.42, 1.55) | 1.52*** (1.43, 1.62) |
| Year | 0.98*** (0.97, 0.98) | 0.98*** (0.97, 0.98) | 0.98*** (0.97, 0.98) | 0.98*** (0.97, 0.98) | 0.99* (0.98, 1.00) | 0.99* (0.98, 1.00) |
| State | ||||||
| Utah (Ref) | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Arizona | 4.13*** (3.35, 5.08) | 4.13*** (3.36, 5.09) | 3.60*** (2.90, 4.46) | 3.61*** (2.92, 4.47) | 1.16 (0.96, 1.40) | 1.16 (0.97, 1.40) |
| Idaho | 2.65*** (2.35, 2.98) | 2.64*** (2.35, 2.97) | 2.55*** (2.28, 2.85) | 2.54*** (2.28, 2.84) | 1.29*** (1.15, 1.44) | 1.29*** (1.15, 1.44) |
| Kansas | 3.30*** (2.92, 3.73) | 3.30*** (2.92, 3.73) | 3.06*** (2.69, 3.48) | 3.06*** (2.69, 3.48) | 1.21* (1.02, 1.43) | 1.21* (1.02, 1.43) |
| Michigan | 3.87*** (3.19, 4.70) | 3.88*** (3.21, 4.70) | 3.29*** (2.70, 4.01) | 3.31*** (2.72, 4.02) | 1.09 (0.91, 1.31) | 1.09 (0.91, 1.31) |
| Missouri | 3.86*** (3.25, 4.59) | 3.88*** (3.27, 4.61) | 3.57*** (3.00, 4.24) | 3.58*** (3.01, 4.25) | 1.31** (1.08, 1.59) | 1.32** (1.09, 1.59) |
| Nebraska | 3.62*** (3.19, 4.11) | 3.62*** (3.20, 4.11) | 3.26*** (2.86, 3.71) | 3.27*** (2.87, 3.72) | 1.26** (1.08, 1.47) | 1.26** (1.08, 1.48) |
| New Hampshire | 3.50*** (2.91, 4.22) | 3.53*** (2.93, 4.25) | 2.81*** (2.35, 3.36) | 2.83*** (2.37, 3.38) | 0.94 (0.78, 1.14) | 0.94 (0.78, 1.15) |
| North Carolina | 3.13*** (2.67, 3.67) | 3.13*** (2.67, 3.66) | 2.46*** (2.08, 2.91) | 2.46*** (2.08, 2.91) | 0.88 (0.73, 1.05) | 0.88 (0.73, 1.05) |
| North Dakota | 4.82*** (4.26, 5.45) | 4.81*** (4.26, 5.44) | 3.83*** (3.42, 4.30) | 3.83*** (3.41, 4.29) | 1.39*** (1.24, 1.57) | 1.39*** (1.24, 1.57) |
| South Carolina | 3.56*** (2.98, 4.25) | 3.56*** (2.98, 4.25) | 2.94*** (2.45, 3.53) | 2.95*** (2.46, 3.53) | 1.02 (0.85, 1.22) | 1.02 (0.85, 1.22) |
| Vermont | 3.52*** (3.08, 4.02) | 3.54*** (3.10, 4.04) | 2.74*** (2.44, 3.09) | 2.76*** (2.45, 3.10) | 0.86* (0.76, 0.98) | 0.86* (0.76, 0.98) |
| Underage drinking policies | ||||||
| None (Ref) | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 | 1.00 |
| Keg registration | 1.07 (0.95, 1.22) | 1.07 (0.95, 1.21) | 1.08 (0.95, 1.23) | 1.08 (0.95, 1.23) | 1.05 (0.92, 1.19) | 1.04 (0.92, 1.19) |
| Social host | 1.01 (0.93, 1.10) | 1.01 (0.93, 1.10) | 0.98 (0.90, 1.07) | 0.98 (0.90, 1.07) | 0.99 (0.90, 1.09) | 0.99 (0.90, 1.09) |
| Consumption | 0.97 (0.86, 1.10) | 0.98 (0.86, 1.11) | 0.96 (0.85, 1.09) | 0.96 (0.85, 1.09) | 0.98 (0.86, 1.11) | 0.98 (0.87, 1.11) |
| IP | 0.90* (0.83, 0.98) | 0.94 (0.87, 1.02) | 1.04 (0.96, 1.13) | |||
| IP interactionsa | ||||||
| IP × ≤ 14 y | 0.85* (0.75, 0.97) | 0.88 (0.75, 1.02) | 1.00 (0.83, 1.21) | |||
| IP × 15 y | 0.89* (0.82, 0.97) | 0.97 (0.89, 1.07) | 1.11 (0.99, 1.25) | |||
| IP × 16 y | 0.96 (0.87, 1.06) | 0.99 (0.89, 1.11) | 1.03 (0.91, 1.16) | |||
| IP × 17 y | 1.02 (0.92, 1.14) | 1.05 (0.93, 1.18) | 1.06 (0.93, 1.20) | |||
| IP × ≥ 18 y | 1.12 (0.99, 1.27) | 1.15* (1.00, 1.32) | 1.15 (0.98, 1.36) | |||
| IP × Male | 0.73*** (0.64, 0.84) | 0.76** (0.65, 0.87) | 0.93 (0.78, 1.12) | |||
| IP × Female | 0.85* (0.75, 0.97) | 0.88 (0.75, 1.02) | 1.00 (0.83, 1.21) | |||
Note. CI = confidence interval; IP = internal possession; OR = odds ratio.
Odds ratio for each group represents the main effect combined with the interaction effect for that group.
*P < .05. **P < .01. ***P < .001.
Model 1 in Table 3 shows that the presence of an IP law was significantly related to a 10% reduction in the odds of past-month drinking. Model 2 (with interactions) reveals that this effect varies across age and gender groups. Although we found bigger reductions in odds among students aged 14 years and younger (15%) and aged 15 years (11%), we found no statistically significant effect among students aged 16, 17, or 18 years and older. Furthermore, the IP law was associated with a larger reduction in the odds of any drinking among male students (27%) than among female students (15%; P < .001 for relative difference; data not shown).
Model 3 indicates that the IP law was associated with a small but not significant reduction (6%) in the odds of past-month binge drinking. Model 4 (with interactions) reveals that the IP law had a significant effect in reducing binge drinking among male students (24%), whereas we found no significant reduction for female students or any of the age groups. In fact, among students aged 18 years or older, the IP law was associated with a significant increase in the odds of past-month binge drinking (15%). However, models of past-month binge drinking among current drinkers (models 5 and 6) found the IP law had no statistically significant effects.
Patterns in student demographics were similar across both drinking behaviors. As expected, older students were more likely to report any drinking and binge drinking than were students in the younger age groups. In general, the odds of each outcome increased with each year of age. Male students had higher odds than female students for both drinking behaviors. Compared with White students, Hispanic students had higher odds for both drinking behaviors, and Black students had much lower odds.
We found none of the other 3 underage drinking policies examined to be significantly related to either of the outcomes. Finally, the odds ratios for the year of survey showed a significant decline in both drinking behaviors over time during the study period (1991–2009).
DISCUSSION
We found that the presence of an IP law is associated with a reduction in the risk of drinking, especially among youths aged 15 years and younger. This finding suggests that IP laws have the intended effect on reducing underage drinking, particularly among younger students. This is especially important to note because by reducing drinking at younger ages, IP laws may help delay the onset of drinking during the developmental stage of adolescence. Research has found that early onset of drinking is associated with increased risk for alcohol abuse and dependence later in life.11–13 It is also associated with other alcohol-related problems, such as drinking 5 or more drinks per occasion, drinking and driving, and injuring oneself or someone else under the influence of alcohol.13 These studies have specifically called for policy interventions targeted at delaying the age of initiation of alcohol use. Moreover, our results indicate that the impact of IP laws was stronger among male students than among female students, which is particularly important because males historically consume alcohol at higher rates than do females.
We included keg registration laws, social host liability laws, and prohibitions against alcohol consumption by minors in each model as control variables. We found none of these policies to be significantly related to either of the outcomes, which does not necessarily mean that these policies had no effect. To evaluate these laws, data selection and analyses would have to be centered at prelaw and postlaw comparisons with respect to each of the specific laws, which was outside the scope of this study. Stratified analysis of only current drinkers—those students who reported any drinking in the past month—revealed that IP laws had no significant effect on binge drinking, thus suggesting that the reduction in the odds of binge drinking among male students observed in the model based on all students was driven by reductions in the prevalence of any drinking. In other words, if male students are less likely to drink at all in the presence of an IP law, then it follows that they are also less likely to binge drink. Conversely, it also appears that if youths are not deterred from drinking by this policy, they are also not deterred from engaging in the risky behavior of binge drinking.
A major strength of this study is the use of state YRBS data, which cover a nearly 20-year period. These surveys have proven reliable and offer consistent results over time.14 The alcohol questions in particular have been shown to be highly reliable with a test–retest agreement of 0.90 for past-month drinking and 0.89 for binge drinking.14
The large sample sizes for the data add credence to the results. These data also displayed patterns consistent with existing research on underage drinking. Our results align with national studies, which have shown that White adolescents drink more and binge drink more than do other racial groups, that male adolescents drink more than do female adolescents, and that the prevalence of drinking among adolescents increases steadily with each year of age.15 We also found that underage drinking rates have declined since 1991, which is consistent with overall drinking trends found in other studies.16 Moreover, as a sensitivity test, we ran similar models with data from all states that have YRBS data available, regardless of whether they have implemented an IP law. The patterns of results are consistent with the results presented in this article (data not shown).
The study is subject to certain limitations. First, the YRBS survey data limit our sample to high school students. Dropouts, younger age groups, and college-aged youths are therefore not included in our analysis. Whether the results would extend to these groups is not clear.
Second, although we included changes in 3 other relevant underage drinking policies in our models in an attempt to isolate the effects of IP laws, other unmeasured contextual forces may be at work. Our variables also did not capture large variability in the underage drinking policies. For example, Utah outlaws kegs entirely, whereas other states require a retailer to collect a name and address from someone purchasing a keg. We coded both Utah and these other states as a “yes” on that law in our analysis. A review by Wagenaar et al.17 documented the wide variation in the statutory and regulatory provisions of keg registration laws among states. Similar variability is found among social host liability laws and, to a lesser extent, consumption laws.
Third, our policy variables were not able to account for varying degrees of stringency or level of enforcement. A study of 11th graders across Oregon found that communities with higher levels of perceived enforcement of underage possession laws had lower rates of drinking and binge drinking.18 The varying levels of enforcement or perceived enforcement across the states and localities in this study may have contributed to the results. However, the inclusion of states as fixed effects should capture the general environment regarding underage drinking within a state and may mitigate this shortcoming.
Moreover, we were unable to include local-level policies in our analysis. A wide variety of policies in place at the local level may affect an underage person’s drinking behavior as much as, if not more than, state-level policies. Finally, most of the IP laws were passed in the late 1990s and early 2000s. Additional years of data are needed to further assess the effectiveness of the IP laws on reducing alcohol-related youth risk behaviors.
Despite these limitations, our analysis highlights IP laws as a potentially valuable policy tool in the fight against underage drinking. Although this is an important first step, further evaluation of this policy is needed as more data (covering longer periods and more states) become available.
Acknowledgments
This article is based on a study conducted for the Alcohol Epidemiologic Data System project funded by the National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health (contract no. HHSN267200800023C) to CSR, Incorporated.
We thank the AJPH reviewers for their comments, which helped to improve the presentation of the study.
Note. The views and opinions expressed in this report are those of the authors and should not be construed to represent the views of the sponsoring agency or the federal government.
Human Participant Protection
The study consisted of secondary data analysis, so institutional review board approval was not necessary.
References
- 1. National Minimum Drinking Age Act, 23 U.S. §158 (1984)
- 2.Wagenaar AC, Toomey TL. Effects of minimum drinking age laws from 1960–2000. J Stud Alcohol Suppl. 2002;63(14):206–225 [DOI] [PubMed] [Google Scholar]
- 3. Welcome to the Alcohol Policy Information System National Institute on Alcohol Abuse and Alcoholism Web site. Available at: http://www.alcoholpolicy.niaaa.nih.gov. Accessed April 1, 2011 [PMC free article] [PubMed]
- 4.Fell JC, Fisher DA, Voas RB, Blackman K, Tippetts AS. The impact of underage drinking laws on alcohol-related fatal crashes of young drivers. Alcohol Clin Exp Res. 2009;33(7):1208–1219 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Fahrenthold D. In N.H., a beer in the belly can get youths arrested. Washington Post. February 5, 2006. Available at: http://www.washingtonpost.com/wp-dyn/content/article/2006/02/04/AR2006020400821.html. Accessed March 25, 2010 [Google Scholar]
- 6.Fell JC, Fisher DA, Voas RB, Blackman K, Tippetts AS. The relationship of underage drinking laws to reductions in drinking drivers in fatal crashes in the United States. Accid Anal Prev. 2008;40(4):1430–1440 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wagenaar AC, Lenk KM, Toomey TL. Policies to reduce underage drinking: a review of the recent literature. In: Galanter M, Ed. Recent Developments in Alcoholism: Alcohol Problems among Adolescents and Young Adults. New York, NY: Plenum Press; 2005: 275–297 [PubMed] [Google Scholar]
- 8.Ringwalt CL, Paschall MJ. The utility of keg registration laws: a cross-sectional study. J Adolesc Health. 2011;48(1):106–108 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Cook MJ. Underage Alcohol Consumption in the United States: Associations with Access Laws, Alcohol Excise Taxes, and Enforcement Practices [master’s thesis]. Storrs, CT: University of Connecticut; 2008 [Google Scholar]
- 10.Cohen DA, Mason K, Scribner R. The population consumption model, alcohol control practices, and alcohol-related traffic fatalities. Prev Med. 2002;34(2):187–197 [DOI] [PubMed] [Google Scholar]
- 11.Dawson DA, Goldstein RB, Chou SP, Ruan WJ, Grant BF. Age at first drink and the first incidence of adult-onset DSM-IV alcohol use disorders. Alcohol Clin Exp Res. 2008;32(12):2149–2160 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Grant BF, Dawson DA. Age at onset of alcohol use and its association with DSM-IV alcohol abuse and dependence: results from the National Longitudinal Alcohol Epidemiologic Survey. J Subst Abuse. 1997;9:103–110 [DOI] [PubMed] [Google Scholar]
- 13.Hingson RW, Zha W. Age of drinking onset, alcohol use disorders, frequent heavy drinking, and unintentionally injuring oneself and others after drinking. Pediatrics. 2009;123(6):1477–1484 [DOI] [PubMed] [Google Scholar]
- 14.Rosenbaum JE. Truth or consequences: the intertemporal consistency of adolescent self-support on the Youth Risk Behavior Survey. Am J Epidemiol. 2009;169(11):1388–1397 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Fisher LB, Miles IW, Austin S, Camargo CA, Colditz G. Predictors of initiation of alcohol use among US adolescents: findings from a prospective cohort study. Arch Pediatr Adolesc Med. 2007;161(10):959–966 [DOI] [PubMed] [Google Scholar]
- 16.Chen C, Yi H, Faden V. Trends in Underage Drinking in the United States, 1991–2009. NIAAA Surveillance Report 91. Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism; 2011 [Google Scholar]
- 17.Wagenaar AC, Harwood EM, Silianoff C, Toomey TL. Measuring public policy: the case of beer keg registration laws. Eval Program Plann. 2005;28(4):359–367 [Google Scholar]
- 18.Dent CW, Grube JW, Biglan A. Community level alcohol availability and enforcement of possession laws as predictors of youth drinking. Prev Med. 2005;40(3):355–362 [DOI] [PMC free article] [PubMed] [Google Scholar]