Regulatory Proposal for Reducing Fire Risks from Cigarettes

Sensitivity and Uncertainty Analysis - Chapter 6

6.1 Introduction

The findings presented thus far in analysing the costs and benefits of a cigarette ignition propensity standard are dependent upon certain key assumptions.  Among the most potentially important of these assumptions are the values employed for two parameters:  (1) the discount rate used to calculate the present value of the regulation's costs and benefits, and (2) the value of statistical life (VSL) estimate used to calculate the economic benefits of reducing cigarette-fire fatalities.

This chapter evaluates the sensitivity of the report's findings to the values employed for these parameters.  This is accomplished by recalculating the present value of costs and benefits using a range of alternate discount rates, and by recalculating benefits using a range of VSL estimates.  Additional detail on each of these sensitivity analyses is presented below.

6.2 Discount Rate Sensitivity Analysis

The economic evaluation of the proposed ignition propensity standard uses a discount rate of three percent to calculate the present value of the regulation's costs and benefits.  This value is within the range of current estimates of the consumption rate of interest and is in keeping with the growing consensus on social discounting methods.⁶⁹  Estimates of the consumption rate of interest, however, remain the subject of some debate, and consensus around use of the consumption rate of interest as the basis for social discounting is not absolute. 

To explore the implications of alternative rates, this analysis presents an evaluation of the present value of costs and benefits using discount rates ranging from one to seven percent, a range that is consistent with that recently employed in analyses of other government regulations.⁷⁰  Consistent with guidance provided by the Treasury Board of Canada Secretariat, the sensitivity analysis also incorporates the use of two higher discount rates:  10 and 15 percent.⁷¹  For purposes of calculating present values, the analysis assumes that both annual costs and annual benefits remain constant in perpetuity.  Exhibits 6-1 and 6-2 present the results.  As the exhibits show, the present value of both costs and benefits diminishes as the discount rate increases.  In all cases, however, the estimated present value of net benefits is positive.

Exhibit 6-1 - Present Value of the Costs and Benefits of an Ignition Propensity Standard (2002 $CAD, millions)

Discount Rate	Costs		Benefits
	Modelled Cost Estimate	Industry-based Cost Estimate	Scenario 1	Scenario 2
1 percent	2,600.0	5,300.0	22,812.2	11,406.4
3 percent	866.7	1,766.7	7,604.1	3,802.1
5 percent	520.0	1,060.0	4,562.4	2,281.3
7 percent	371.4	757.1	3,258.9	1,629.5
10 percent	260.0	530.0	2,281.2	1,140.6
15 percent	173.3	353.3	1,520.8	760.4

Exhibit 6-2 - Present Value of the Net Benefits of an Ignition Propensity Standard (2002 $CAD, millions)

Benefit Scenario	Discount Rate	Cost Scenario
		Modelled Cost Estimate	Industry-based Cost Estimate
Scenario 1	1 Percent	20,212.2	17,512.2
	3 Percent	6,737.4	5,837.4
	5 Percent	4,042.4	3,502.4
	7 Percent	2,887.5	2,501.7
	10 Percent	2,021.2	1,751.2
	15 Percent	1,347.5	1,167.5
Scenario 2	1 Percent	8,806.4	6,106.4
	3 Percent	2,935.5	2,035.5
	5 Percent	1,761.3	1,221.3
	7 Percent	1,258.1	872.3
	10 Percent	880.6	610.6
	15 Percent	587.1	407.1

6.3 Value of Statistical Life Sensitivity Analysis

As discussed in Chapter 5, the value of statistical life (VSL) approach allows analysts to estimate the economic value of health and safety initiatives that, on a statistical basis, are expected to save lives.  A variety of economic techniques can be employed to derive VSL estimates, and VSL estimates presented in the literature vary widely.⁷²  As a result, estimates of VSL values are considered uncertain.  The analysis presented in Chapter 5 relies on a VSL estimate of $5.8 million (2002 $CAD), the central VSL estimate derived as part of an extensive literature review presented in Air Quality Valuation Model Version 3.0 (AQVM 3.0) Report 2: Methodology.⁷³  Exhibit 6-3 presents the annual and present value of benefits of the proposed regulations using the central, high, and low VSL estimates presented in the AQVM 3.0 Report.  As the exhibit shows, use of the low VSL figure would reduce estimated annual benefits under Scenario 1 from $228.1 million to $145.3 million, while use of the high VSL figure would increase estimated annual benefits to $440.5 million.  Under Scenario 2, use of the low VSL figure would reduce the annual benefits estimate from $114.1 million to $72.7 million; in contrast, use of the high VSL figure would raise the annual benefits estimate to $220.3 million.

Exhibit 6-3 - Sensitivity of Benefits to Value of Statistical Life Estimates (2002 $CAD)

Benefit Scenario	VSL Estimate	Value1	Annual Benefits2	Present Value of Benefits3
Scenario 1	Low	$3.5 million	$145.3 million	$4,844.1 million
	Central	$5.8 million	$228.1 million	$7,604.1 million
	High	$11.7 million	$440.5 million	$14,684.1 million
Scenario 2	Low	$3.5 million	$72.7 million	$2,422.1 million
	Central	$5.8 million	$114.1 million	$3,802.1 million
	High	$11.7 million	$220.3 million	$7,342.1 million

Notes:
¹ Adjusted from 1996 to 2002 Canadian dollars using the Canadian Medical Cost Index for Healthcare.
² Includes the benefits of a reduction in cigarette-fire fatalities, injuries, and property damage.
³ Based on a discount rate of three percent.

---

⁶⁹ For a review of the literature on social discounting, see: U.S. Environmental Protection Agency, Guidelines for Preparing Economic Analyses, September 2000, Chapter 6.

⁷⁰ See, for example, Health Canada , 2003, Benefit-Cost Analysis of Alternative Maximum Allowable Concentration Guidelines for TTHMs in Canadian Drinking Water, August 12.

⁷¹ This guidance is provided in the Treasury Board's draft Benefit Cost Analysis Guide, which is dated July 1998 and is available on-line at http://www.tbs-sct.gc.ca/fin/sigs/Revolving_Funds/bcag/BCA2_E.asp. This guidance replaces the previous Treasury Board Benefit Cost Analysis Guide, which was published in 1976.

⁷² See, for instance, Fisher, A., L. Chestnut, and D. Violette (1989), "The Value of Reducing Risks of Death: A Note On New Evidence," Journal of Policy Analysis and Management 8(1): 88-100; Miller, T. (1990), "The Plausible Range for the Value of Life - Red Herrings Among the Mackerel," Journal of Forensic Economics 3(3): 17-39; and Viscusi, W.K. (1993), "The Value of Risks to Life and Health," Journal of Economic Literature, December: 1912-1946.

⁷³ L.G. Chestnut, D. Mills, and Stratus Consulting, Air Quality Valuation Model Version 3.0 (AQVM 3.0), Report 2: Methodology, September 3, 1999.