Transcript Outcome Evaluation of Paid Counter

Impact of Tobacco Control Program
Expenditures on Tobacco Use
Maya Vijayaraghavan, Ph.D.
Office on Smoking and Health
National Conference on Tobacco or Health
November 19-21, 2002
TM
Outline of Presentation
• Purpose of the studies
• Overview
• Sources of data
• Review of two studies:
– Expenditures and Aggregate Cigarette Sales
– Expenditures and Individual Cigarette Use
• Implications for policy
TM
Purpose of the studies
Do changes in Tobacco Control Program (TCP)
expenditures influence the
– Prevalence of smoking
– Consumption of cigarettes
independent of other factors that influence
the demand for cigarettes
TM
Overview
• Aggregate-level analysis: impact of TCP
expenditures on aggregate cigarette
consumption
• Individual-level analysis: impact of TCP
expenditures on the
– prevalence
– intensity of smoking
TM
Sources of Data
Policy Variables
• TCP Expenditures
–
–
–
–
CDC’s IMPACT
NCI’s ASSIST
RWJ Smokeless States
State TCP
• Cigarette Excise Taxes
– The Tax Burden on Tobacco (The Tobacco Institute, 1998)
TM
Sources of Data
Other variables
• Tax-paid Cigarette Sales
– The Tax Burden on Tobacco (The Tobacco Institute, 1998)
• Population and Demographics
– United States Census Bureau
• Aggregate income
– United States Bureau of Economic Analysis
• Individual-level data
– Behavioral Risk Factor Surveillance System
(BRFSS)
TM
TCP Expenditures and Aggregate
Cigarette Sales (1981-1998)
Methods:
• Data: cigarette sales and TCP expenditures
over time for all the states
• Controlled for cigarette excise taxes, crossborder cigarette sales, demographics and
income
• Multivariate econometric analysis
TM
TCP Expenditures and Aggregate
Cigarette Sales (1981-1998)
Results:
• Consistent evidence that TCPs have decreased
aggregate consumption of cigarettes
• Effect in states with comprehensive TCPs
(Arizona, California, Massachusetts, Oregon)
was 10-50 times greater than for the US overall
TM
TCP Expenditures and Aggregate
Cigarette Sales (1981-1998)
Results:
• Programs that secure  $1 in per capita
funding had lower per capita sales than those
< $1
• Programs funded above $6 per capita may
result in greater declines in sales
• At $1 to $6 funding levels, the effect was
larger one year after reaching this level of
funding
TM
TCP Expenditures and Aggregate
Cigarette Sales (1981-1998)
Results:
• Increases in expenditures had their greatest
impact when tobacco use was relatively high
• Increases in excise taxes had the same impact
regardless of level of cigarette consumption
• States with higher taxes than neighboring
states had fewer tax-paid sales
TM
TCP Expenditures and Individual
Cigarette Use (1985-1998)
Methods:
• Data : smoking prevalence, quantity smoked,
and TCP expenditures over time
• Controlled for cigarette excise taxes, income,
demographics
• Multivariate econometric analysis to estimate
prevalence and consumption
TM
TCP Expenditures and Individual
Cigarette Use (1985-1998)
Results:
• Increases in expenditures - declines in
smoking prevalence and number of cigarettes
smoked
• Past expenditures continued to contribute to
declines for more than a year
• More smokers quit rather than reducing the
quantity smoked
TM
TCP Expenditures and Individual
Cigarette Use (1985-1998)
Race/Ethnicity:
• TCPs most effective in reducing smoking
among Hispanics and whites
• For all ethnicities, program effects declined
with age
• Increases in cigarette prices effective in
reducing quantity smoked among Whites only
• No impact found for African-Americans
TM
TCP Expenditures and Individual
Cigarette Use (1985-1998)
Age:
• Older smokers less likely to change
behavior due to increase in price
• TCPs most effective among 18-24 year
olds
• Impact declined with age
TM
TCP Expenditures and Individual
Cigarette Use (1985-1998)
Gender:
• TCPs more successful at curbing
smoking among women of all ages than
among men
• Women of all ages more price sensitive
than men
TM
Implications for Policy
• TCPs work!
• TCPs are more effective, dollar for dollar,
when implemented on a larger scale
• Appropriate funding levels lead to reduced
consumption of cigarettes
• Lag effect of TCP expenditures means
funding levels must be maintained
TM
Implications for Policy
• Regardless of prevalence, increasing excise
taxes leads to reduced consumption of
cigarettes
• Focus on prevention and early intervention
may be more cost-effective than targeting
older smokers
• Past approaches may have to be revisited to
ensure that they are appropriately tailored to
various populations
TM
References
Farrelly M, Pechacek T, Chaloupka F, 2001. The Impact
of Tobacco Control Program Expenditures on Aggregate Cigarette
Sales: 1981-1998; National Bureau of Economic Research: Working
Paper 8691.
Farrelly M, Nimsch C, Cowell A, 2001. Coordinated
Evaluation of Statewide Tobacco Control Programs in the US,
Phase Two: Impact of Tobacco Control Program Expenditures on
Individual Tobacco Use—1985-1998 Behavioral Risk Factor
Surveillance Systems. Unpublished draft report.
TM
CDC
Office on Smoking and Health
www.cdc.gov/tobacco
770.488.5309
TM