Contents
Cost-benefit analysis, also called benefit-cost analysis, is a method of comparing fire safety options in which the consequences of an investment are valued in money and divided into benefits and costs.[1] The amounts are converted to present worths or annual equivalents, so that options whose costs and benefits fall at different times can be compared.[1] In fire protection, the benefits are mainly the lives saved, injuries avoided and property damage reduced by a measure, and the costs are its initial and ongoing costs.[2] Hall and Watts (2008) call it the most common approach to risk evaluation, the step in which a calculated risk reduction is judged.[3] A variant, cost-effectiveness analysis, expresses the benefits on a single non-monetary scale, such as the cost per life saved.[2]
A single option is acceptable when its benefits exceed its costs. The best of several options is the one at which the marginal benefit of further protection stops exceeding the marginal cost.[4] The method is used for building designs, community fire protection plans and new regulations.[5][6] It requires parameters that are controversial, notably the value of a reduced risk of death and the discount rate used to compare future and present consequences.[2]
Definition and place in fire risk analysis
Watts and Chapman define the method in the Handbook of Fire Protection Engineering of the Society of Fire Protection Engineers (SFPE, third edition, 2002). They describe benefit-cost analysis, also referred to as cost-benefit analysis, as a comparison in which the consequences of an investment are evaluated in monetary terms.[1] The consequences are divided into benefits and costs, which are then converted to annual equivalents or present worths. The analysis has four steps: identifying the relevant benefits and costs, measuring them, selecting the best alternative and treating uncertainty.[1]
Hall and Watts treat the method in the Fire Protection Handbook of the National Fire Protection Association (NFPA, 2008). They present it as a technique of risk evaluation in which all benefits of risk reduction are translated into monetary equivalents.[3] A new product or suppression system can then be assessed by its net profit or loss, by its total cost plus loss, or by a ratio. Here "profit" means saved lives, avoided injuries and reduced property damage on one monetary scale, and "loss" means the initial and ongoing cost of the new product or system.[2] A fire risk analysis that includes such an evaluation step may be called a fire risk assessment.[3]
The risk analysis supplies the benefit side. Watts and Hall (2002) describe cost-benefit analysis as a specific type of decision analysis, in which a fire risk analysis provides estimates for some of the benefits and another analysis quantifies the costs.[7] In the framework of Hall and Watts (2008), fire scenarios, probabilities and consequences give the expected losses per year with and without a proposed change. A loss evaluation model expresses the change in money, and a cost-benefit comparison model sets it against the annual cost of the change to give a measure of its attractiveness.[8] In a quantitative risk assessment for the chemical process industries, Barry (2002) describes relative risk reduction assessment as a cost-benefit tool for finding the best balance between prevention, protection and emergency response.[9]
Watts and Chapman (2002) note that economic analysis in fire protection engineering is mainly concerned with cost reduction. Typical questions are the least expensive way to meet a requirement, or how to minimise the sum of expected fire losses and the investment in fire protection.[10] They list four common methods of comparing alternative investments: present worth, annual cost, rate of return and benefit-cost analysis.[10]
Identifying and measuring benefits and costs
For fire prevention or control, the benefits are based on the fire losses that occur without the activity.[1] Fire loss is divided into direct losses, which are property and contents losses, and indirect losses. Indirect losses include the costs of injuries and deaths, business interruption, interrupted community services, lost payroll or taxes, lost market share and damage to reputation. Direct costs of fire protection include fire-resistive construction, the installation of protection systems and the operation of fire departments.[1]
Which benefits count depends on who decides. For a property owner, the relevant benefits are likely to be lower fire insurance premiums and smaller damage or business interruption losses that insurance does not cover.[1] For a municipality, they include protection of the community, avoided tax and payroll losses and the costs of assisting fire victims. A community may still ignore effects outside its borders. The 1953 fire in an automobile transmission plant in Livonia, Michigan, affected the car industry in Detroit and dealers throughout the United States, but the community had little incentive to count those losses.[1] Watts and Chapman suggest that the wider the decision unit, the more likely it is that all relevant costs and benefits are included.[1]
Direct losses are measured from actuarial fire loss data or by experienced judgement of the maximum probable loss.[1] Indirect losses are sometimes estimated as a percentage or multiple of direct losses. An NFPA study of 109 fires from 1989 added 65 per cent for manufacturing and industrial properties and 25 per cent for public assembly, educational, institutional, retail and office properties.[11] It added 10 per cent for residential, storage and special structures, and nothing for vehicle and outdoor fires.[11] Costs are valued as opportunity costs: the cost of a fire department is the other community services that the same resources could have provided.[1]
At community level, Granito (2008) describes the total cost of fire as the fire loss plus the costs of prevention, suppression and fire insurance.[12] Calculating it allows a fire department and its community to estimate the cost-effectiveness of a proposed prevention programme.[12] Recorded each year and adjusted for inflation and growth, it also lets a community compare its current fire performance with earlier years.[13] At national level, the World Fire Statistics Centre tracks four core cost components: fire damage, fire-fighting organisations, extra construction costs for fire safety and insurance administration.[11] For the United States from 1980 to 1997, these core components typically amounted to about 1 per cent of gross domestic product.[11]
The benefit of a measure also depends on whether it works when needed. Hall and Watts (2008) note that a manager can order a sprinkler system but must rely on others to install it correctly, and that its reliability decreases without scheduled testing and maintenance.[3] Analyses therefore need probabilities for the status of each system, such as fully operational, blocked or switched off. In this way the reliability of a fire protection system affects the estimated benefit.[3] The International Fire Engineering Guidelines (IFEG, 2005) include the annual costs of inspecting and maintaining fire equipment on the cost side of their fire cost expectation.[14]
Discounting and the time value of money
Costs and benefits rarely occur at the same time. Watts and Chapman (2002) give three reasons why a sum of $1,000 now is worth more than the same sum one year later: inflation, risk and the cost of money.[15] A future sum is therefore converted to a present worth by discounting at an interest rate per period, and a uniform annual amount is converted with the series present worth factor.[16] The inverse, the capital recovery factor, spreads an initial investment over its life as an equivalent annual cost.[16]
As an example from the source, $1,000 received in 5 years at 10 per cent has a present worth of $620.92.[16] The choice of discount rate matters because the cost of a fire-resistive building is incurred early while its benefits accrue over the life of the building.[10] The higher the discount rate, the lower the present value of future benefits, so analysts often test a range of rates.[10] In their example, the ranking of two projects even switches at a discount rate of 15 per cent.[10] Watts and Chapman also warn that inflated prices must be combined with a nominal discount rate and constant prices with a real rate; mixing them causes serious distortions.[17]
Their worked example of the Beverly Hills Supper Club fire shows the effect of discounting.[17] An analysis compared annual savings of $11,000 in insurance premiums with sprinkler installation costs of $42,000 to $68,000. It concluded that the sprinklers would have paid for themselves in four to seven years. That analysis treated $11,000 received after four years as equal to $11,000 in the first year. With future savings discounted at 10 per cent, the payback period becomes five to eleven years.[17]
A second example compares a partial and a full sprinkler system at 10 per cent interest.[17] The partial system costs $8,000, lasts 15 years and leaves an annual insurance premium of $1,000. The full system costs $15,000, lasts 20 years and leaves a premium of $250. The annual costs are $1,000 + $8,000 × 0.13147, or about $2,052, for the partial system and $250 + $15,000 × 0.11746, or about $2,012, for the full system. The full system is slightly more economical, and when costs are this close, other criteria such as uninsured losses become especially important.[17]
Decision criteria
Watts and Chapman (2002) separate two questions: whether a single project is acceptable, and which of several projects to select.[4] A project is acceptable if its benefits minus its costs exceed zero, or if its benefit-cost ratio exceeds one.[4] Both criteria use the present worth or annual equivalent of each side.[1]
For selection, the degree of fire protection is increased step by step. Marginal costs are low at first and then rise, while the marginal benefit curve is less well known and may be flat. The economically optimum level of fire protection lies where the two curves cross; beyond it, the extra safety costs more than it yields.[4] The source illustrates this with a numerical example of five strategies, A to E, with costs and losses in dollars.
| Strategy | A | B | C | D | E |
|---|---|---|---|---|---|
| Fire reduction costs | 0 | 10 | 25 | 45 | 70 |
| Fire losses | 100 | 70 | 50 | 40 | 35 |
| Sum of costs and losses | 100 | 80 | 75 | 85 | 105 |
| Marginal benefit | – | 30 | 20 | 10 | 5 |
| Marginal cost | – | 10 | 15 | 20 | 25 |
| Marginal benefit-cost ratio | – | 3.0 | 1.33 | 0.5 | 0.2 |
The highest marginal ratio occurs at strategy B, yet strategy C is the optimum, because the step from B to C still yields an extra net benefit of $5.[4] The same answer follows from the sum of fire losses and fire reduction costs, which is the life-cycle cost of each strategy. It falls to a minimum of $75 at C and then rises. As an illustration of why the marginal view matters, strategy D still has a benefit-cost ratio of $60/$45 = 1.33 measured against A, although its last step returns only half its cost.
Uncertainty can be introduced by treating losses as random variables. With probabilities for no fire, a minor, an intermediate and a major fire, each with its own damage, the expected loss is the probability-weighted sum.[18] Each protection strategy then changes some probabilities or damages and adds its own cost, and the strategy with the lowest expected loss is preferred.[18]
The IFEG (2005) use the same idea in their fire cost expectation (FCE), which they call a present value. Their diagram adds the capital cost of active and passive fire protection, annual inspection and maintenance costs, and expected building and contents fire losses.[14] The diagram does not show how the annual amounts are discounted. To combine these terms as a present value, the recurring costs and expected losses must first be put on the same present-value basis as the capital cost. The equations below summarise the expected-loss calculation and, schematically, the FCE components.
| Symbol | Description | Unit |
|---|---|---|
| Present worth of a sum of money | $ | |
| Future sum of money | $ | |
| Uniform end-of-period amount (annual cost or saving when the period is a year) | $/period | |
| Interest or discount rate per period | – | |
| Number of periods | periods | |
| , | Present worth (or annual equivalent) of benefits and of costs | $ (or $/period) |
| Benefit-cost ratio | – | |
| Probability of fire outcome i (no, minor, intermediate or major fire) | – | |
| Damage loss for outcome i | $ | |
| Expected loss | $ (per period) | |
| , , | Capital cost of protection, present value of inspection and maintenance costs, present value of expected fire losses | $ |
| Value per statistical life implied by willingness to pay | $ per life | |
| Amount a person is willing to pay for a risk reduction | $ | |
| Reduction in the probability of death | – |
Valuing life and injury
Hall and Watts (2008) name the value of a reduced risk of death or injury as one of the controversial parameters of the method.[2] Because of this, risk evaluation sometimes uses methods that downplay such parameters, including the acceptable risk approach, which they describe as a kind of backdoor cost-benefit analysis.[2] Ramachandran (2002b) stresses that a dollar value for loss of life is not a price for losing one's life. It is meant to reflect a social consensus on the value of changes in the risk of death by fire.[19]
He gives a simple example of how such a value arises. If most people would pay $1,500 to reduce their lifetime risk of dying in a fire from one chance in 500 to one chance in 1,000, the risk falls by 1/1,000.[19] They then value a life saved at $1,500 per 1/1,000 of a life, or $1.5 million per life. Four approaches to valuing life have been identified: lost output, life insurance amounts, court awards and willingness to pay. The last is the most widely adopted, and the acceptable expenditure per life saved is likely to be higher for involuntary than for voluntary risks.[19]
The estimates vary widely. A 1981 study cited values of statistical life from $50,000 to $8 million.[20] The NFPA studies of the total cost of fire in the United States used $5 million per death and $166,000 per injury as 1993 values. These values follow the practice of economists at the US Consumer Product Safety Commission. On that basis, the dollar equivalent of US fire deaths and injuries was $42.5 billion in 1997.[20]
Cost-effectiveness measures avoid the explicit value. Notarianni (2002) describes a municipal model of mandating residential sprinklers in the United States, where approximately 3,500 people died each year in residential fires at the time.[21] Its outcome criteria were dollars per premature death averted and dollars per life-year saved, so no value of life had to be specified. Using 1989–1993 fire data, the net cost was lowest, at $1.4 million per death averted, for a new mobile home in a small community in the South. It was highest, at $35.1 million, for a retrofit in a one- or two-family dwelling in a medium-size community in the West. The national average calculation gave a median of $7.3 million.[21]
Applications
Cost-benefit analysis is used to set regulations. The IFEG (2005) explain that the fire-related Performance Requirements of the Building Code of Australia provide a benchmark risk that takes account of community expectations and of the cost to the community. That cost may be determined by a cost benefit analysis, which may be necessary for new requirements.[6] In New Zealand, according to the same guidelines, Regulatory Impact Statements and cost benefit analyses had to be completed for all new requirements before they could enter the New Zealand Building Code.[22] The ABCB handbook (2020) adds that changes to the Australian deemed-to-satisfy provisions since 1988 were often supported by fire safety analyses and cost benefit analyses.[23]
In his 2008 account of United States practice, Schultz notes that NFPA 101, Life Safety Code, reflected cost-benefit considerations in its sprinkler requirements for mercantile occupancies. Its height and area limits then left many small shops unsprinklered, subject to an exception for accessory uses, while requiring sprinklers in medium and large shops.[24] Cote and Grant (2008) observe that cost-benefit models may be used to justify an increased level of performance in a code, or to show that a suggested level is impractical economically.[25] They also point out that each further increment of safety typically costs more as the risk becomes lower.[26]
Cost-benefit studies have also been used to assess product and vehicle measures. In the United Kingdom, regulations of the late 1980s required upholstered furniture and mattresses, apart from mattress tickings, to resist small flames and cigarettes. An official UK survey concluded that the improved materials raised costs by $24 to $32 per furniture item and gave a benefit-cost ratio of 40:1.[27] A US statistician's reanalysis did not support that conclusion.[27] For a water mist system in aircraft passenger compartments, developed from 1985 by aviation authorities and industry, a cost-benefit study found the cost of outfitting fleets too high relative to the lives saved. The system was therefore not made mandatory.[28]
Communities apply the method to their fire protection plans. Granito (2008) writes that a fire protection master plan should seek the maximum cost-benefit ratio for fire protection expenditures and set up a way of measuring their effectiveness.[5] Citizens ultimately decide the trade-offs on the level of protection they buy, but the safety of emergency responders cannot be reduced as part of a cost-reduction programme.[13] For public fire safety education, Hall (2008) describes dividing a community into areas and estimating a loss value for each, such as its average fires or fire deaths over five years. Discounted by the expected success of the programme, this is the benefit, which is divided by the cost of delivery. The resulting ratios rank the areas, and a line is drawn below the last area that fits the budget.[29]
Limitations and criticism
Hall and Watts (2008) identify the value of reduced death or injury risk and the discount rate as controversial inputs.[2] Watts and Chapman (2002) show that the latter can change the ranking of investments and advise testing a range of rates.[10] Cote and Grant (2008) add that no solely economic, engineering or public health approach can do justice to all the social, economic, political and legal factors that shape how people value safety.[25]
Expected values have limits too. Watts and Chapman (2002) note that an expected loss reflects only the average of the probability distribution.[18] Ramachandran (2002b) shows the consequence with insurance. By the expected value principle, an owner should insure only if the premium is less than the expected loss, while the insurer offers cover only if the premium is higher. The risk aversion of most people resolves this: they prefer a small fixed loss to the risk of a large one.[30]
Average inputs can also hide large differences. Municipal decisions on mandatory systems are often based on national average values, and uncertainty is rarely considered.[31] In the residential sprinkler model, the cost for new mobile homes was as low as one fifth of the national average result, while for existing single-family homes it was up to five times higher.[32]
Finally, an analysis is exposed to the interests of those who commission it. Van Rossum (2027) warns that clients can gain from cheaper construction methods, a lower level of installed protection, dropping costly measures or cheaper materials. Such incentives may tempt engineers to act unethically.[33] Where an assessing authority cannot easily check a complex analysis, it must be able to rely on the integrity of the engineer, because a second-opinion review is not economically feasible for every project.[34]
See also
Notes
- ↑ 1.00 1.01 1.02 1.03 1.04 1.05 1.06 1.07 1.08 1.09 1.10 1.11 Watts and Chapman (2002), p. 5-99.
- ↑ 2.0 2.1 2.2 2.3 2.4 2.5 2.6 Hall and Watts (2008), p. 3-138.
- ↑ 3.0 3.1 3.2 3.3 3.4 Hall and Watts (2008), p. 3-137.
- ↑ 4.0 4.1 4.2 4.3 4.4 Watts and Chapman (2002), pp. 5-99–5-100.
- ↑ 5.0 5.1 Granito (2008), p. 12-20.
- ↑ 6.0 6.1 National Research Council of Canada et al. (2005), Part 0 – Australia, p. 0.2-3.
- ↑ Watts and Hall (2002), p. 5-6.
- ↑ Hall and Watts (2008), pp. 3-139–3-141.
- ↑ Barry (2002), p. 5-176.
- ↑ 10.0 10.1 10.2 10.3 10.4 10.5 Watts and Chapman (2002), p. 5-97.
- ↑ 11.0 11.1 11.2 11.3 Ramachandran (2002b), pp. 5-79–5-80.
- ↑ 12.0 12.1 Granito (2008), p. 12-15.
- ↑ 13.0 13.1 Granito (2008), p. 12-7.
- ↑ 14.0 14.1 National Research Council of Canada et al. (2005), p. 2.3-5.
- ↑ Watts and Chapman (2002), p. 5-93.
- ↑ 16.0 16.1 16.2 Watts and Chapman (2002), pp. 5-95–5-96.
- ↑ 17.0 17.1 17.2 17.3 17.4 Watts and Chapman (2002), p. 5-98.
- ↑ 18.0 18.1 18.2 Watts and Chapman (2002), pp. 5-100–5-101.
- ↑ 19.0 19.1 19.2 Ramachandran (2002b), p. 5-85.
- ↑ 20.0 20.1 Ramachandran (2002b), p. 5-86.
- ↑ 21.0 21.1 Notarianni (2002), pp. 5-61–5-62.
- ↑ National Research Council of Canada et al. (2005), Part 0 – New Zealand, p. 0.2-5.
- ↑ Australian Building Codes Board (2020b), p. 14.
- ↑ Schultz (2008), p. 20-99.
- ↑ 25.0 25.1 Cote and Grant (2008), p. 1-57.
- ↑ Cote and Grant (2008), p. 1-56.
- ↑ 27.0 27.1 Babrauskas (2008b), p. 6-107.
- ↑ Mawhinney (2008), p. 16-141.
- ↑ Hall (2008), pp. 5-114–5-115.
- ↑ Ramachandran (2002b), pp. 5-87–5-88.
- ↑ Notarianni (2002), p. 5-60.
- ↑ Notarianni (2002), p. 5-63.
- ↑ Van Rossum (2027), §2 and §2.2.
- ↑ Van Rossum (2027), §2.1.
References
- Australian Building Codes Board. (2020b). Handbook: Fire Safety Verification Method (Print version 1.1).
- Babrauskas, V. (2008b). Upholstered furniture and mattresses. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 6-103–6-128). National Fire Protection Association.
- Barry, T. F. (2002). Quantitative risk assessment in chemical process industries. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 5-176–5-213). National Fire Protection Association; Society of Fire Protection Engineers.
- Cote, A. E., & Grant, C. C. (2008). Codes and standards for the built environment. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 1-51–1-71). National Fire Protection Association.
- Granito, J. A. (2008). Planning for public fire-rescue protection. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 12-3–12-22). National Fire Protection Association.
- Hall, J. R., Jr. (2008). Using data for public education planning and decision making. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 5-103–5-118). National Fire Protection Association.
- Hall, J. R., Jr., & Watts, J. M., Jr. (2008). Fire risk analysis. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 3-135–3-143). National Fire Protection Association.
- Mawhinney, J. R. (2008). Water mist fire suppression systems. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 16-135–16-180). National Fire Protection Association.
- National Research Council of Canada, International Code Council, Department of Building and Housing, & Australian Building Codes Board. (2005). International fire engineering guidelines (2005 ed.). Australian Building Codes Board.
- Notarianni, K. A. (2002). Uncertainty. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 5-40–5-64). National Fire Protection Association; Society of Fire Protection Engineers.
- Ramachandran, G. (2002b). Measuring fire consequences in economic terms (J. R. Hall Jr., Rev.). In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 5-79–5-92). National Fire Protection Association; Society of Fire Protection Engineers.
- Schultz, E. (2008). Mercantile occupancies. In A. E. Cote, C. C. Grant, J. R. Hall, Jr., & R. E. Solomon (Eds.), Fire protection handbook (20th ed., pp. 20-97–20-103). National Fire Protection Association.
- Van Rossum, E. (2027). Introduction to Fire Safety Engineering. Fire Safety Academy.
- Watts, J. M., Jr., & Chapman, R. E. (2002). Engineering economics. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 5-93–5-104). National Fire Protection Association; Society of Fire Protection Engineers.
- Watts, J. M., Jr., & Hall, J. R., Jr. (2002). Introduction to fire risk analysis. In P. J. DiNenno, D. Drysdale, C. L. Beyler, W. D. Walton, R. L. P. Custer, J. R. Hall Jr., & J. M. Watts Jr. (Eds.), SFPE handbook of fire protection engineering (3rd ed., pp. 5-1–5-7). National Fire Protection Association; Society of Fire Protection Engineers.