University of Washington professor of social work Richard Catalano and colleagues studied whether anti-alcohol policies in public and private schools in Washington state and Australia's Victoria state were effective for eighth- and ninth-graders.

What they found was that each school's particular policy mattered less than the students' perceived enforcement of it. So, even if a school had a suspension or expulsion policy, if students felt the school didn't enforce it then they were more likely to drink on campus. But, even if a school's policy was less harsh -- such as requiring counseling -- students were less likely to drink at school if they believed school officials would enforce it.

"Whatever your school policy is, lax enforcement is related to more drinking," Catalano said.

The study was published recently in the journal Health Education Research.

The results were similar in Washington, where the legal drinking age is 21 and schools tend to have a zero-tolerance approach, and Victoria, Australia, where the legal drinking age is 18 and policies are more about minimizing harm.

In the study, 44 percent of Victoria eighth-graders and 22 percent of Washington eighth-graders reported drinking alcohol. Victoria students also reported higher rates of binge drinking and alcohol-related harms.

Apart from perceptions about enforcement, harmful behaviors in both states were reduced when students believed policy violators would likely be counseled by a teacher on the dangers of alcohol use, rather than expelled or suspended.

"Schools should focus on zero tolerance and abstinence in primary and early middle school, but sometime between middle school and high school they have to blend in zero tolerance with harm minimization," said Catalano, director of the Social Development Research Group at the UW School of Social Work and principal investigator for the International Youth Development Study. "By the time they get into high school they need new strategies."

Those strategies could include talking to a teacher or being referred to treatment. The likelihood of binge drinking was reduced if students received an abstinence alcohol message or a harm minimization message, and if they believed teachers would talk to them about the dangers of alcohol. Catalano said such remediation policies are an important predictor of less alcohol use among ninth-graders.

He said the study shows harsh punishment for drinking on school grounds, such as calling the police or expelling the student, doesn't inhibit alcohol use on campus. Instead, long-term negative impacts of expulsion mean students feel disconnected from school and may subsequently drink more. Calling the police, which gives the student a police record, appears to make things even worse.

"What we've seen in other studies from this sample is suspension policies actually worsen the behavior problem," Catalano said. "What that says to me is, although you want policies and you want enforcement of policies, there are other ways of responding than suspension, expulsion and calling the police: Getting a student to talk to a teacher about how alcohol might be harmful, or a session with the school counselor."

The study was funded by the National Institute on Drug Abuse, National Institute on Alcoholism and Alcohol Abuse, and Victorian Government's Operational Infrastructure Support Program. Co-authors are Todd Herrenkohl of the UW, lead author Tracy Evans-Whipp and Stephanie Plenty of Royal Children's Hospital in Victoria, Australia, and John Toumbourou of Deakin University in Australia.