South African/SACU Experience with
Tobacco Taxation: Lessons Learned and
Relevance for other SADC countries
Corne van Walbeek
Associate Professor, School of Economics
University of Cape Town
Presentation to a World Bank-sponsored workshop on tobacco taxation and illicit
trade in Gaborone, June 2012
1
Tobacco use and taxation has a long
history
• King James I of England (1604):
“….a custom loathsome to the eye, hateful to the nose,
harmful to the brain, dangerous to the lungs, and in the
black stinking fume thereof nearest resembling the
horrible Stygian smoke of the pit that is bottomless.”
• Adam Smith (1776):
“sugar, rum, and tobacco, are commodities which are no
where necessaries of life, which are become objects of
almost universal consumption, and which are therefore
extremely proper subjects of taxation.”
2
Until 1994 the tobacco industry and the
government were strongly aligned
Minister of Finance, 1983: “The Tobacco Board has presented
justified arguments for the maintenance of the status quo
regarding the excise taxes on tobacco, and I do not intend to
wake sleeping dogs”
Minister of Finance, 1986: “any increases in excise duties at
present could be counter-productive, since it could in fact –
on account of the potentially adverse effect on consumption
– lead to a reduction of revenue from this source”
Article 5.3 of the FCTC:
In setting and implementing their public health policies with
respect to tobacco control, Parties shall act to protect these
policies from commercial and other vested interests of the
tobacco industry in accordance with national law
3
Excise tax policy in SA since 1994
• In 1994 the government announced that it intended to raise the excise tax
to 50% of the retail price
– Explicit acknowledgement that it had a public health goal
• Subsequently this statement was reinterpreted to mean that total tax
burden (i.e. excise tax plus VAT should equal 50% of the retail price)
• In 2006 the total excise tax target was raised to 52% of the retail price
• Result of this excise tax policy:
– Treasury passively changes the specific excise tax to maintain the 50% (or 52%)
tax burden
– High degree of predictability for both Treasury and industry
– Government has largely transferred tobacco control policy to the industry,
away from itself
4
The South African experience, 1961 - 2011
6000
500
5000
400
4000
300
3000
200
2000
100
1000
Between
1993 and
2011 real
excise tax
increases by
487% and real
excise tax
revenue
increases by
249%
Real excise rate
Real excise revenue
2009
2005
2001
1997
1993
1989
1985
1981
1977
1973
1969
0
1965
0
Real excise revenue
(R million, 2000 prices)
600
1961
Real excise rate
(in constant 2000 cents)
Big increases in the excise tax have resulted in big increases in tax
revenue
5
Why are excise tax increases so effective in
reducing tobacco consumption?
• Despite its addictiveness, people respond to changes in the price of
cigarettes
• Price elasticity of demand for cigarettes in SA (and many low- and
middle-income countries is around -0.6)
• Thus a 10% increase in cigarette prices reduces consumption by 6%
• Not all people respond to changes in cigarette prices, but the poor
and youth are typically more responsive than the better-off and the
old
• Some indicative numbers:
– 1000 smokers initially
– If price increases by 10%, we would expect 3% (= 30) smokers to quit
– Remaining smokers will cut down their consumption by an average of
3% (some more and some less)
6
Setting the excise tax in SA
• Tobacco excise tax and other “sin taxes” are levied as specific
tax
• Specific taxes have some nice features:
– Easy to determine the quantum of tax
– Auditing is much easier
– Differences between high-price and low-price cigarettes are smaller
than with ad valorem excise tax
– BUT specific tax is eroded by inflation if not adjusted regularly
• Excise tax is set by Treasury
• Announced by Minister of Finance at reading of the budget
• Taxes apply to all SACU members (SA, Botswana, Namibia,
Lesotho and Swaziland)
– Other SACU members may charge higher excise taxes (e.g. Botswana
on alcohol)
7
2008/2009
2006/2007
2004/2005
2002/2003
2000/2001
1998/1999
1996/1997
1994/1995
1992/1993
1990/1991
1988/1989
1986/1987
1984/1985
1982/1983
1980/1981
1978/1979
1976/1977
1974/1975
1972/1973
1970/1971
1968/1969
2.00
1966/1967
3.00
1964/1965
1962/1963
1960/1961
Excise tax, Rand per 20 cigarettes, nominal values
7.00
160
6.00
140
5.00
Nominal excise tax per pack of 20
120
Consumer Price Index
100
4.00
80
Government commits to a TC policy
based on keeping the tax burden at
50% of the retail price
60
40
1.00
20
0.00
0
Consumer Price Index (2010 = 100)
Since 1994 rapid excise tax increases followed the small
nominal increases in the 1970s and 1980s
9.00
200
8.00
180
8
10.00
9.00
8.00
7.00
6.00
5.00
4.00
3.00
2.00
1.00
0.00
1945/1946
1948/1949
1951/1952
1954/1955
1957/1958
1960/1961
1963/1964
1966/1967
1969/1970
1972/1973
1975/1976
1978/1979
1981/1982
1984/1985
1987/1988
1990/1991
1993/1994
1996/1997
1999/2000
2002/2003
2005/2006
2008/2009
2011/2012
Excise tax, Rand per 10 cigarettes, constant
2010 prices
Lesson learnt
• In an inflationary environment and/or in a long-term
comparative situation, one should focus on real values, rather
than nominal values
Real excise tax on cigarettes
9
The industry’s pricing strategy and its impact
• BATSA is a near-monopolist with at least 92% share of (legal)
sales
• Being a monopolist it has significant pricing power
• In a “dark market” it is difficult for competitive firms to enter
the market
• BATSA is a price leader; other firms follow
• Understanding price elasticity, the industry knows that it can
raise its revenues by increasing the net-of-tax price of
cigarettes
• Since the government is already raising the retail price by
increasing the excise tax (thus reducing consumption and
industry revenues), the industry responds by raising the netof-tax price
10
The quantum of the excise tax is determined by
the net-of-tax price
11
How does this work in practice?
Government raises excise tax
Retail price increases
Consumption decreases
Industry revenue decreases (if net-of-tax price does
not change)
Industry increases the net-of-tax price
Retail price increases
Government raises excise tax to maintain 52% tax
burden
12
25.00
20.00
15.00
10.00
5.00
Net-of-tax price
Excise tax
2008
2006
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
1976
1974
1972
0.00
1970
Retail price (constant 2008 prices)
A long-term view on the composition of
the retail price
GST/VAT
13
Despite a large fall in consumption, real excise revenues
and industry revenues have greatly increased
14000
2000
12000
10000
1500
8000
1000
6000
4000
500
2000
Cigarette consumption (RHS scale)
Excise tax revenue
2008
2006
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
1976
1974
0
1972
0
Aggregate cigarette consumption (millions of packs)
2500
1970
R millions (adjusted for inflation, 2008 base)
16000
Total industry revenue
14
Percentage changes in important variables
since 1993
Variable
Percentage
Real excise tax per pack of cigarettes
378%
Real net-of-tax price of cigarettes
153%
Real retail price
212%
Cigarette consumption
-33%
Per capita cigarette consumption
-51%
Real excise tax revenue
220%
Real industry revenue
69%
Smoking prevalence
From about 35% to about 22%
15
Aggregate consumption
2,500
90
80
2,000
70
60
1,500
50
1,000
40
30
500
20
10
0
0
Per capita consumption (pack per person
aged 15 +)
2011/2012
2008/2009
2005/2006
2002/2003
1999/2000
1996/1997
1993/1994
1990/1991
1987/1988
1984/1985
1981/1982
1978/1979
1975/1976
1972/1973
1969/1970
1966/1967
1963/1964
1960/1961
1957/1958
1954/1955
1951/1952
1948/1949
1945/1946
Aggregate consumption (million packs)
Cigarette consumption in SA, 1946 - 2011
Per capita consumption (age 15 +)
16
Some important issues and lessons
1. The ex ante (stated) burden of the tax and the ex
post (actual) burden of the tax
2. Cigarette affordability
3. The threat of illicit trade
17
The actual burden of the tax vs. the stated
burden of the tax
• Stated total tax burden = 52% of the retail price of the “most
popular brand” category
• This seems to be the recommended retail price, as published
by BATSA
• Actual average selling price of most popular brand is higher
than recommended selling price, based on Stats SA survey of
prices
• It seems that the tax increase announced at budget time does
not take into consideration that the tax is passed onto
consumers in form of higher selling price
• Result: the actual tax burden always runs behind the targeted
tax burden
18
Because of the pricing and excise tax dynamics,
the stated tax burden is not reached
Source: Unpublished data on prices, Statistics South Africa
19
A possible solution
• When calculating the tax increase required to maintain the
52% tax burden, assume that the tax will be fully passed on to
the consumer
• Adopt a completely new excise tax model, not based on a
targeted tax burden, but where the quantum of the excise tax
increase is determined exogenously of the industry’s pricing
mechanism
20
Is cigarette affordability important?
• Cigarette tax and price are important determinants of
cigarette consumption
• Income growth is also an important determinant of
consumption
– As average income levels increase, cigarette consumption increases
• Affordability of anything depends on the price and income
• Affordability index for cigarettes: what percentage of per
capita GDP is required to buy 100 packs of cigarettes?
– If this index increases, more resources are required to buy a pack of
cigarettes, and thus cigarettes become less affordable (and vice versa)
21
90
4.50
80
4.00
70
3.50
60
3.00
50
2.50
40
2.00
30
1.50
Cigarette affordability
2008
2006
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
0
1980
0.00
1978
10
1976
0.50
1974
20
1972
1.00
Per capita cigarette consumption (packs/year)
(population aged 15+)
5.00
1970
Percentage of per capita GDP required to buy 100
packs of cigarettes
An uncannily close relationship between cigarette
consumption and affordability (r = - 0.98)
Per capita consumption (RHS scale)
22
Implications of this
• When average incomes increase, cigarette consumption tends
to increase, ceteris paribus
• We want rapid economic growth, but presumably we do not
want rapid increases in cigarette consumption
• Increases in the excise tax (and price) of cigarettes should
counteract the impact of increases in average income
• Suggestion: Increase the excise tax, taking cognizance of both
the inflation rate and changes in real GDP
23
The threat of smuggling and other illicit
activity
• Industry argues that illicit trade is big and growing
• Of course there is illicit cigarette activity in South
Africa
• Economic theory indicates that greater excise taxes
create incentives for tax evasion and avoidance
• The contention is:
– How big is the problem?
– How much of an impact does it have on the crucial
parameters (especially government revenue and
consumption)?
24
The scale of the illicit trade problem
2006: TISA “illicit trade has grown from zero ten years ago to 20% of the
total South African market”
2007-2010: TISA estimates the illicit market share as 20% of the market
2011: TISA estimates the illicit market as 25% of the South African market
Year
Illicit sales in
billion sticks
(BAT numbers)
Legal sales in
billion sticks
(BAT numbers)
Implied total
sales in billion
sticks
(BAT numbers)
Illicit market
share
(percentage)
Legal sales in
billion sticks
(official
numbers)
2008
3.1
25.1
28.2
11.0
25.3
2009
4.6
23.1
27.7
16.6
24.2
2010
6.3
21.0
27.3
23.1
21.6
Source: BAT, 2010
25
The problem
• We cannot easily get believable data on the
magnitude of the share of illicit cigarettes, without
doing comprehensive and expensive surveys
• The question briefly addressed here:
– Can we determine, from official sources, whether cigarette
consumption has decreased to the extent that it becomes
a significant problem for government revenue?
26
If illicit trade in tobacco becomes an
increasing problem, one would expect the
following:
– Large deviations in actual revenue from
budgeted revenue
– Consistent over-budgeting of revenues (and
thus disappointing actual revenues)
27
How large are the deviations from budget, on
average?
Theil’s U-coefficient: zero = no deviation; large U-coefficient = large deviation, on average
Financial years by
decade
Beer made from
malt
Spirits
Cigarettes and
tobacco products
1950/51 to 1959/60
0.090
0.090
0.221
1960/61 to 1969/70
0.133
0.049
0.040
1970/71 to 1979/80
0.093
0.140
0.047
1980/81 to 1989/90
0.084
0.115
0.034
1990/91 to 1999/2000
0.058
0.087
0.083
2000/01 to 2011/12
0.066
0.112
0.049
28
15%
10%
5%
0%
-5%
-10%
-15%
-20%
Cigarettes and tobacco
Spirits
2010/2011
2008/2009
2006/2007
2004/2005
2002/2003
2000/2001
1998/1999
1996/1997
1994/1995
1992/1993
-25%
1990/1991
Actual revenue less budgeted revenue, as
a percentage of actual revenue
There is no evidence that the Treasury officials were
consistently over-budgeting tobacco revenue
Average deviation from
budget (1994/5 to 2011/12):
Beer: - 1.7%
Spirits: - 4.9%
Cigarettes & tobacco: + 0.3%
Beer (made from malt)
The budgeted revenues for cigarette and tobacco excise taxes were very accurate
29
on average; for beer and especially spirits they were rather optimistic
Conclusion
• South Africa’s excise tax model has worked well in the past
but the quantum of the excise tax increase is wholly
controlled by the tobacco industry
• A tax model where the excise tax increases in a predictable
amount in real terms, and accounts for the growth in real
GDP, is required
• There is little evidence to indicate that the growth in illicit
trade has significantly undermined government’s excise
revenue
• The decrease in the quantity of cigarettes consumed is in line
with what one would have expected, given the tax and price
increases of the past two decades
30