Dr. David Merriman: Cigarette Tax Avoidance and Evasion - Evidence from Littered Cigarette Packs
1. Cigarette tax avoidance and evasion:
Evidence from littered cigarette packs
by David Merriman
Professor of Public Administration, University of Illinois at
Chicago and Visiting Fellow, Lincoln Institute of Land Policy
Remarks prepared for:
Drexel School of Economics and Econsult Solutions, Inc.
2nd Annual Urban Economic Policy Conference: Topics in Urban Tax Policy
Philadelphia, Pennsylvania
Wednesday, April 27,2016
1
4. Tax stamps on littered
cigarette packs generally
indicate the state (and
sometimes local gov’t) of
purchase
4
5. • a small army of
enthusiastic
undergraduates collected
more than 1,000 littered
cigarette packs
• 135 randomly selected
areas chosen to provide a
representative sample of
the population of cigarettes
smoked in Chicago
• Summer of 2007.
5
7. Probability of an Indiana stamp fell rapidly
as distance to Indiana border rose
A one mile increase in distance to the lower tax state border increased
the probability of a local stamp by approximately one percent
7
8. Explanations other than tax avoidance
• Incidental stamp mismatches due to tourism and commuting
• Tourism is mostly in the Chicago loop area. Ignoring packs collected in this area
results in roughly similar results.
• Commuting (very unlikely to matter much)
• A reasonable assumption is that in the absence of tax differentials
• people who work in Chicago (but live elsewhere) would purchase 1/3rd of packs in the city and
smoke 1/3 in the city.
• people working in the suburbs but living in the city would purchase 2/3rds of packs in the and
smoke 2/3rds in the city
• Putting these assumptions together with Census data on commuting suggests that in the
absence of tax differentials about 99% of littered packs found in Chicago would have an
Illinois tax stamp.
• Litters are not representative because they are unconventional people
who take joy in breaking rules.
8
9. Representativeness of littered packs?
Evidence from a small sample of public garbage receptacles in the same location as litter collections found no
evidence of higher levels of tax avoidance among littered packs. 9
10. Two minutes on some other littered pack
studies
•Five east coast cities (Davis, Grimshaw,
Merriman, et. al. 2013)
•NYC (Chernick and Merriman 2013)
10
12. After NYS $1.25 tax increase
NYC (Chernick and Merriman 2013)
Avoidance (measured
with littered packs)
went up after $1.25
increase in NYS
cigarette tax
12
Note that our estimates suggest that
after the tax increase
1. Tax revenue increased
2. Consumption decreased about
10% even after we take into
account illicit sales
15. some thoughts about implications for soda
taxes
15
20oz bottle
No tax
$1.50
With 3 cent per oz tax
$2.10
No tax
$6.44
With $3.66 per pack tax
$10.00
6 pack of 20 oz
No tax
$6.50 (?)
With 3 cent per oz tax
$6.50+$3.00=$9.50
(1) (2) (3)
Is the appropriate analogy
(1) versus (2)? Or
(1) Versus (3) ?
16. some thoughts about implications for soda
taxes (continued)
16
Suppose (1) (cigarette pack) versus (3) (6 pack)
than analogy with cigarette tax avoidance in Chicago and Indiana works pretty well
This suggests:
1. Very high avoidance ( on the order of 80%) on the border of the tax change.
2. Very steep gradient—avoidance will decline rapidly with distance. Perhaps one
percent per mile.
3. “How wide (effectively) is the river?” is an important question. (In NYC we found
that avoidance from NJ was very small despite $1.67 difference in tax)
4. Consumption is likely to drop substantially despite high level of avoidance
17. some thoughts about implications for soda
taxes (continued)
17
Suppose (1) (cigarette pack) versus (2) (20 ounce bottle)
than analogy with cigarette tax avoidance in Chicago and Indiana suggests:
It probably won’t be worth the bother for many people to avoid especially if
1. People like their soda cold and don’t have easy access to quick refrigeration
2. Soda purchases are more transitory and impulsive than purchases of a pack
of 20 cigarettes
18. some (very preliminary) thoughts about
implications for soda taxes (continued)
18
Bottom line: analogies with cigarette tax have to be very tentative but…
1. Ultimate result is likely to depend upon many geographic and context specific
variables.
2. There would probably be
A. significant avoidance and evasion.
B. a significant decline in consumption.
C. a significant amount of revenue compared to sales.