Indoor Smoking Ban

Banning Smoking in Public Places
Cost benefit analysis of the proposed
legislation for England
Public Expenditure Analysis
3rd May 2006
Caroline Godkin
Jennifer Potter
Jennifer Sleppy
1
Structure of this presentation
1. Background and outline of the CBA
2. Review of the costs
3. Review of the Benefits
4. Sensitivity Analysis
5. Summary and Conclusions
2
1.
Background
3
An analysis and critique of the case for a
smoking ban in enclosed public places


Legislation in February 2006 was debated for banning
smoking in all enclosed public places in England
Cost benefit cases were constructed for four options:





To continue with the voluntary approach
To make all enclosed spaces smoke free
To give local authorities powers to make enclosed spaces smoke
free
To make all enclosed spaces smoke free with exceptions
The option that was passed, and that will be considered
in this critique, was a ban in all spaces without
exceptions
4
Some smoking facts about the UK
Percent of Great Britain Population who Smoke
1974 - 2003

50
Percent of Population
45

40
Smoking rates in Great Britain
have fallen every year and are
currently at about 26% of the
population
Cigarettes are extremely highly
taxed
Duty of 22% of price plus £100
per 100
Value added tax of 17.5%
There are significant rates of
cigarette smuggling from
Continental Europe exploiting
lower taxes in other EU countries

35

30

25
20
1974
1982
1990
1994
1998
2001
2003
5
The Net Benefits for a complete ban were
estimated to be c. £1.7bn for a single year
Benefits
Averted deaths from SHS
Averted death from smokers
giving up (quitters)
Year 1
£m
Costs
371
1780
550
Estimated savings to NHS for
reduction in overall prevalence
100
Production gains from reduced
exposure to SHS
Safety benefits
Estimated annual savings for
maintenance and cleaning
TOTAL BENEFITS
Implementation
Enforcement
0
20
Education / Communication
Averted deaths from reduced
uptake of smoking
Reduced sickness absences
Year 1
£m
70-140
340-680
63
100
3,374-3,784
1
Tax revenue losses from falling
tobacco sales
Employees
1,145
Customers
150
Losses to the tobacco industry
129
Production losses (smoking breaks)
430
Consumer surplus losses to
continuing smokers
155
TOTAL COSTS
2,030
6
Smokers, smokers trying to quit and
non-smokers are all given standing




The study considered the losses in consumer surplus for
smokers who would not be able to smoke indoors both at
work and in bars
The consumer surplus loss of those trying to quit was not
considered as it was viewed that they would be “happy
to give it up”
The considerable number of non-smokers were given
standing through the reduction to second hand smoke
The tobacco industry as well as The Exchequer were
given standing through consideration of their losses from
any ban
7
The study contained few details of
calculations and was inconsistent



The study used many calculations and
assumptions that were not stated in the paper
Detailed research was required to be able to
understand values given for costs and
benefits
Both sections were full of inconsistencies


Reductions in smokers and amount of cigarettes
smoked varied
Items were double counted
8
2.
Review of the costs
9
The total costs are estimated to be
£2bn for the first year
Costs
Implementation
Enforcement
Education / Communication
Year 1
£m
0
20
1
Tax revenue losses from falling tobacco sales
Employees
1,145
Customers
150
Losses to the tobacco industry
129
Production losses (smoking breaks)
430
Consumer surplus losses to continuing smokers
155
TOTAL COSTS
These are
the costs
that will be
looked at in
detail
2,030
10
Many of the costs are overstated



For employees the loss in their consumer surplus is
counted as though they would stop smoking
altogether – you would just go outside to smoke
In addition counting the production losses from
people going outside double counts the employee
losses i.e. you either stop smoking or you go outside
Losses to the tobacco industry are taken at a flat
10% of the tax revenue with no explanation of the
basis of the calculation
11
Inconsistent assumptions are used


Employee smokers’ consumer surplus is
considered but not that of customers
Overall the predicted fall in revenue
represents a 19% fall in duty for England



In 2004 – 05 tobacco duty collected in the UK was
£8.1m or £6.88m for England
The suggested revenue losses for the smoking
ban is £1.3m or 19% of the total revenue receipts
Compare this with reduction in smoking of 6.5%
12
The cost to the health service of
additional years of life is not considered



Increasing life expectancy will place an
additional burden both on the health service
and also on the pensions system
The report states that there has been no
additional value assigned to the costs of
increasing health care and pensions
We have considered the additional costs of
healthcare after 10 years of the ban being in
place
13
After the first year we have assumed
the costs will reduce



Revenue losses to the Exchequer will reduce
as there are fewer people giving up smoking
Consumer surplus losses will be reduced as
there are fewer people smoking
We have assumed that the additional health
and pension costs will be seen after 10 years
of the ban
14
Our costs are significantly lower than
the costs estimated in the study
Sources of Decreased Costs
2,500
2,030
1,024
Pounds Million
2,000
1,500
118
63
48
Decrease in
losses to the
tobacco
industry
Increase in
production
losses
Decrease in
Consumer
surplus losses
1,000
904
500
0
Orignal Study Decrease in tax
losses
Our
Assessment
15
3.
Review of Benefits
16
The study miscalculated the benefits
Health Benefits:

Reduction in illness and mortality from:
Lung cancer
Heart disease
Asthma attacks
Childhood respiratory disease
Sudden infant death syndrome

Life expectancy gain from reduced smoking uptake as a result of the ban

Life expectancy gain for smokers who quit as a result of the ban





Environmental and Economic Benefits:

Reduction in insurance costs to the Department of Health, National Health
Services (NHS), from reduced smoking and exposure to secondhand
smoke

Reduced cost from sickness absences and productivity gains

Reduced fire risk – including damage, injury, and death

Reduced cleaning and maintenance costs
17
Research from other studies informed
the calculations for our study

Department of Transport






Value of a human life lost at 43 yrs - £1m
Value of an injury at 43 yrs - £40,000
Value of an additional year of life £28,571
Scaling the research to reflect the population
of England
Estimates of costs to the NHS for smokers
British medical journal - death rates from
SHS in public places
18
Estimate of Benefits from Smoking
Ban
Benefits from RIA study
Averted deaths from SHS
Year 1
£m
371
Averted death from smokers giving up (quitters)
1780
Averted deaths from reduced uptake of smoking
550
Estimated savings to NHS for reduction in overall prevalence
100
Reduced sickness absences
Production gains from reduced exposure to SHS
Safety benefits
Estimated annual savings for maintenance and cleaning
TOTAL BENEFITS
70-140
340-680
63
100
3,374-3,784
19
Our benefits are significantly higher than
the benefits estimated in the study
Sources of Increased Benefits
726
4,500
322
4,250
40
140
3,875
Decrease in
estimated
savings to
NHS
Decreased
deduced
sickness
absences
Our Estimate
Pounds Million
4,000
3,750
3,614
42
Orignal
Study
Increase in
averted
deaths from
SHS
3,500
3,250
3,000
2,750
2,500
Increase in
averted
death from
smokers
giving up
Decrease in
averted
deaths from
reduced
uptake of
smoking
20
4.
Sensitivity analysis
21
Original study ignored sensitivity; we
tested our assumptions













Discount Rate
Original study did not discount
Base Case: 6% (Treasury Green Book - standard in UK)
Vary Between: O% and 15%
Time Horizon
Original study looked at only 1 year
Base Case: 10 years
Vary Between: 1 year and 40 years
Other Point Estimates
Reduction in cigarettes smoked and people smoking
them
Savings from reduced illness; productivity gains
Vary each around our best estimates
Monte Carlo analysis
22
NPV falls but remains positive as
discount rate increases
NPV Sensitivity to Discount Rate
25,000,000,000
20,000,000,000
15,000,000,000
10,000,000,000
5,000,000,000
9%
10
%
11
%
12
%
13
%
14
%
15
%
8%
7%
6%
5%
4%
3%
2%
1%
0
0%
10 year NPV (Pounds)
30,000,000,000
Discount Rate (%)
23
NPV increases as time horizon
extends
NPV Sensitivity to Time Horizon
40,000,000,000
35,000,000,000
25,000,000,000
20,000,000,000
15,000,000,000
10,000,000,000
5,000,000,000
40
37
34
31
28
25
22
19
16
13
10
7
4
0
1
NPV (Pounds)
30,000,000,000
Number of Years
24
Focus on time benchmarks for easier
assessment
Description
Years
NPV
Original Study
1
£2,971 Million
Max Parliament Term
5
£10,885 Million
Base Case
10
£18,594 Million
Generation View
40
£35,914 Million
25
Used excel’s RAND function to set up
Monte Carlo analysis
Point Estimates
First Year Reduction in amount of cigarette smoked
First Year Reduction in smokers
Expected
High
Low
(RAND()*Range)+Low
8.00%
(RAND()*Range)+Low
1.40%
Range
0%
8.00%
0%
1.40%
Further Year reduction in amount of cigarettes smoked
Further year reduction in smokers
(RAND()*Range)+Low
(RAND()*Range)+Low
0%
0%
Savings from reduced smoking illness and productivity gains
(RAND()*Range)+Low 680,000,000 340,000,000 340,000,000
Discount Rates
(RAND()*Range)+Low
Ran 5,000 iterations
Recorded NPVs
in data table
2.00%
1.00%
12%
0%
2.00%
1.00%
12.00%
Descriptive Statistics
Mean
£18,773 Million
Median
£18,200 Million
Max
£41,369 Million
Min
£3,950 Million
26
Monte Carlo analysis shows NPV will
most likely fall between £15-30B
Histogram of Monte Carlo Analysis for NPV of Smoking Ban
1200
800
600
400
200
M
or
e
0
£5
£0
,0
00
,0
00
£1
,0
00
0,
00
0,
00
0,
£1
00
5,
0
00
0,
00
0,
£2
00
0,
0
00
0,
00
0,
£2
00
5,
0
00
0,
00
0,
£3
00
0,
0
00
0,
00
0,
£3
00
5,
0
00
0,
00
0,
£4
00
0,
0
00
0,
00
0,
£4
00
5,
0
00
0,
00
0,
00
0
Frequency
1000
Bin
27
5.
Summary and conclusions
28
Original study rife with bad data and
opaque analysis

Cost and benefit calculations were:





Not transparent
Based on inconsistent assumptions
Subject to double counting in numerous instances
Failed entirely to consider cost to the health
service of additional years of life
No discounting; involved time horizon of only
one year
29
Shortcomings of study possibly
explained by political realities

Some of the gap may be due to the fact the intent
was to pass a less stringent restriction




Wanted Option 4:
partial ban – smoking allowed where no food served
Got Option 2:
complete ban – all enclosed spaces smoke free
Overstating of costs to the tobacco industry could
reflect significant campaign contributions
Despite the poor analysis, Option 2 was passed into
legislation, and the ban will be in place in 2007.
30