1
Look, Im underage ... Sold!
Tobacco Sales Increasewhen Real ID Is Shown

• Arnold H. Levinson, Ph.D.
• University of Colorado
• Stacy Hendershott
• Colorado Tobacco Enforcement Unit

2
Compliance CheckSurvey

• Six CO urban-suburban counties
• Census of outlets (n1083)
• State-issued ID with true age
• Same minors varied use of ID
• One attempt per outlet

3
Comparison of Conditions,ID vs. No ID

• Balanced
• Male Age
• Clerk gender, perceived age Day of week
• Clerk asked for ID

Unbalanced No ID ID

Female 36 64
Median Time 1130 a.m. 1245 p.m.

4

• Overall sales rate
• 12.7
• Asked to see ID
• 87.2
• Relative risk (RR)
• if ID not asked
• 7.1
• (plt0.0001)

5
Sales Rates,ID Used vs. Not Used (ID Always
Requested)

• Relative risk of sale when ID used
• 6.2
• (plt0.0001)

6
of Attempts Ending in Sale,By MinorCases
Where ID Was Requested
Age Gender No ID Carried ID
17 M 0.0 21.7
15 M 3.4 14.8
16-17 M 20.0 28.6
16 F 0.0 3.7
15 M 0.0 8.7
16 M 0.0 36.8
15 M 0.0 8.3
16 M 0.0 13.0
15 M 2.8 18.8
14 M 2.1 2.7
14 F 0.0 0.0
17 M 6.0 0.0
7
What if

• Logistic model of sale
• with varied ID condition
• Significant covariates
• minor gender
• ages 14 17
• clerk age lt30
• Weekdays
• two counties
• asked for ID

8
Whats Going On?

• Possible explanations
• Too busy to do it right?
• Cant calculate?
• Think showing ID implies 18?
• Think law requires just showing ID?
• Ask just for appearances?

9
Implications for Compliance Checks

• Carry ID raise rates
• (if clerks usually ask).
• Dont use ID lower rates.
• Should we be replicating
• real purchase practices?
• How often do kids use ID
• for real attempts.

10
The Compliance-Use GapReal or artifact?
11
The Issue

• Fewer sales ? less smoking

12
Clerk failure to ask IDof underage cigarette
customers(same counties, time period)
Real (n590)
Simulated (n1083)
13
Sales Rates,Real vs. Simulated Tries
14
Conclusions

• Real underage shoppers asked for ID less than
  compliance checkers.
• Interpretations
• Different looks and actions.
• Real underage shoppers
• dont choose stores randomly.

15

• A few outlets and circumstances
• can meet adolescent demand.
• To Do
• Target stores routinely selling underage.
• Target proxy-buying and social giving.
• Evaluate impact of comprehensive access control
  on adolescent smoking.