The purpose of cost-benefit analysis is to inform decisions taken on behalf of others. It is used to evaluate the provision of public goods - for which consumer preferences are not directly reflected in market prices. Its preferred methodology evaluates those preferences from estimates of their willingness to pay to receive benefits or to avoid disbenefits.
Although its findings may be controversial, approximate, and open to objections on theoretical grounds, it is generally accepted that cost-benefit analysis can provide decision-makers with a better guide than would otherwise be available.
The purpose of cost-benefit analysis is to indicate whether a proposed decision would be consistent with the preferences of those affected. The actions that are required for that purpose are:
- the specification of the scope of the analysis and of the decision criterion to be used;
- the forecasting of the material consequences of the proposed decision;
- the monetary valuation of each consequence in light of the preferences of those that would be affected;
- the aggregation of those valuations in a way that is consistent with the decision criterion; and,
- the interpretation of the aggregated valuations.
The scope of an analysis is normally determined by the requirements of its prospective users. The evaluation of a medical procedure might, for example, be concerned solely with its effects on patients and their families, or it might also encompass effects on the hospital or on the public at large. An evaluation of migration commissioned by a European government might not be required to include effects upon migrants from other European countries, but an evaluation commissioned by the European Union would be expected to require the inclusion of effects upon all of its inhabitants. Some users require the coverage of a longer timescale than others, and environmental studies may even be required to encompass effects on future generations.
The customary ("consequentialist") decision criterion of cost-benefit analysis uses the summation of the effects of the expected consequences of a decision on the well-being of people within its scope. The alternative ("deontologist") decision criterion takes account of predetermined rules that reflect the beliefs or customs of the community concerned. To make a summation possible, a monetary cost or benefit is assigned to each of those effects. The summation normally used is then that of the costs and benefits of each effect, weighted by their estimated probability of occurrence, and discounted using the estimated time preference rates of those affected. The normal acceptance criterion for a single project is a positive net present expected value of the excess of benefits over costs, and the normal criterion for the acceptance of one among several alternatives is that it should be the one with the highest net present expected value. Some authorities specify a required rate of return criterion, but that has the disadvantage that the associated internal rate of return calculation can sometimes produce inconsistent results .
The practice of cost-benefit analysis depends upon the assumption that - although the acceptability of the consequences a decision is determined by the reactions of those affected - the forecasting of those consequences is the responsibility of the decision-maker. That responsibility is assumed to be delegated to a representative government by its voters (and also, for example, to a medical practitioner by his patients). It is here assumed that forecasts of all of the expected outcomes and their probabilities are already available.
The outcomes whose valuations typically figure in cost-benefit analysis, include non-financial benefits such as the saving of time and the relief of suffering, and non-financial costs in the form of unwelcome experiences such as injury and exposure to noise. The monetary value to a person of such an outcome is taken to be the amount that he would be willing to pay in order to enjoy its benefits or avoid its costs. The preferred way of estimating the relevant willingness to pay is the revealed willingness to pay (or revealed preference) method, which depends upon observations of the subjects' conduct in making free and well-informed choices. Valuation can be reliably estimated from the price that is determined by choices made in an efficient market, a reasonable approximation to which is often provided by organised markets for products and services. An alternative is to use a stated preference method  such as a contingent valuation or a multi-attribute valuation, which employ survey questionnaires to establish the subjects' stated willingness to pay. Another possibility is the valuation of particular outcomes according to the observed community willingness to pay in expression of its collective attitude to similar outcomes. (Under that approach, the value of life might be derived from the costs of communally-determined practices concerning its preservation, such as prolonged hospitalisation).
The preferred method of aggregation is to calculate the sum of the net present expected values of all the outcomes, using their estimated probabilities and the appropriate discount rate. An exception is the practice (adopted in some assessments of global warming) of assessing the value of a "worst case scenario" when probability estimates are not available.
Health and safety
- See also Cost-benefit analysis (health care)
The value of life
Estimates of the value of a life have featured in cost-benefit analyses of medical programmes, road safety schemes and health and safety regulations. The evidence on which they are based has come mainly from studies of the market tradeoffs between money and fatality risks, and mainly from the analysis of the premiums paid to employees to compensate them for undertaking risky work. Such risk premium studies have been shown to imply values of life ranging from less than a million dollars to over 20 million (at 2000 prices), rising with income and falling with age. It seems unlikely that contingent valuation estimates exhibit less scatter.
The available evidence thus leaves the users of cost-benefit analysis with a wide range of choice. The valuations that have been adopted include:-
- United States Department of Transportation. Value of a statistical life: $5.8 million @ 2007 prices (range $3.2 to $8.4 million).
- European Commission. Value of a death from environmental pollution (age adjusted): €1 million @ 2000 prices (range €0.65 to €2.5 million) .
- United Kingdom Department for Transport. Average value of prevention per fatal casualty: £1.43 million @ 2005 prices (human cost £0.94 million, lost output £0.49 million)
Unlike the transport and environmental authorities, the medical authorities are able to make informed estimates of the duration of the lives saved and of their likely quality of life. Consequently their evaluations have generally used measures of "quality-adjusted life-years" saved, rather than numbers of lives saved. The United Kingdom's National Institute for Health and Clinical Excellence (NICE) authorities use that measure to stipulate arbitrary  cost-effectiveness thresholds (in the range £20,000 to £30,000 per quality adjusted life-year) rather than performing explicit cost-benefit analyses. Similarly, it does not appear that cost-benefit analysis is applied to medical procedures in the United States. A 2006 analysis of current practice there has concluded that the medical cost per year of life gained had increased from $7,400 in the 1970s to $36,300 in the 1990s . Survey-based quality of life assessments   that are available from international organisations  are in general use, however.
Costs of injuries
The United States and United Kingdom authorites allow for the cost of road traffic injuries by applying injury severity factors to their cost of life estimates that allow for the lower human costs and higher medical costs
Assessments of the costs of premature mortality and morbidity associated with contact with a variety of pollutants are available from the regulatory authorities in the United States, the United Kingdom and Australia. A comparison of the methodology and findings of eight cost-benefit studies provides information obtained from risk-compensation data, contingent valuations and cohort studies.
The preferred method of estimating the cost of noise is by econometric analysis of its effect upon house values, but some attempts have been made to use contingent valuations. Comparisons of substantial numbers of international studies  suggest that nothing better than a broad indication of the orders of magnitude involved has been achieved.
The cost-benefit assessment of climate change poses issues that have not arisen elsewhwere. Firstly, the uncertainty about the material consequences of the alternative decisions is so great that a risk-oriented criterion has often been used, rather than the conventional use of the probability-weighted net present expected value. Secondly, the fact that most of its adverse consequences affect unborn future generations raises ethical questions concerning the choice of discount rate. The Stern Review of the economics of climate change  argued that, on ethical grounds, there should be no difference between the values attributed to a benefit (or cost) now, and the provision of the same benefit (or cost) to someone living at some remote time in the future. The justification for that assumption is discussed below.
Objections and rebuttals
In principle, the impossibility of looking into the minds of others impedes the task of acting in accordance with their preferences. It is an obstacle to the aggregation of effects on their welfare, especially when some benefit and others are harmed. In practice, however, common sense often enables a decision maker to overcome that obstacle. It makes it obvious, for example, that an aggregate gain in economic welfare would be achieved by a road improvement that would save a hundred lives at the expense of requiring the same number of families to move house. It is the task of cost-benefit analysis to provide an objective basis for such decisions by applying the criteria of welfare economics to such information as can be discovered concerning the preferences of those who would be affected. Unfortunately, the criteria that are useful for that purpose are open to criticism. The Kaldor-Hicks criterion, that requires that gainers be able to compensate losers, does not guarantee an increase in aggregate welfare unless the compensation is actually paid, and that requirement is seldom fully met.
A criticism that can be levied against a strictly utilitarian interpretation of cost-benefit analysis is that it is cannot applied consistently, but often has to be overruled. It is considered necessary, for example, to exclude the benefits of crime to its perpetrators. Political philosophers such as John Rawls advocate departures in the interests of fairness, and few believe that, for example, a small benefit to a large majority justifies the imposition of great suffering on a small minority. A possible response to the accusations of paternalism expressed in the course of the intergeneration transfer controversy would be a claim that there is now general agreement that the interests of remote generations should be given equal weight with those of current generations.
There have also been criticisms of cost-benefit analysis on the grounds that the margins of error in most evaluations is so wide as to justify regarding the process as completely unreliable. In its defence it can be credited with reasonable reliability when it yields a present value that is beyond its range of uncertainty.
- Gaylon E. Greer and Phillip T. Kolbe: Investment analysis for real estate decisions (Google books extract), Dearborn Real Estate, 2003
- [http://www.econ.upf.edu/docs/papers/downloads/705.pdf Anna Merino-Castelló: Eliciting Consumers Preferences Using Stated Preference Discrete Choice Models: Contingent Ranking versus Choice Experiment, June, 2003},
- Robert Mitchell and Richard Carson: The Contingent Valuation Method, (From: "Using Surveys to Value Public Goods"), The World Bank Group, 1989
- W Kip Viscusi and Joseph Aldi: The Value of a Statistical Life: A Critical Review of Market Estimates Throughout the World, Harvard Law and Economics Discussion Paper No. 392, November 2002 (free SSRN download)
- Treatment of the Economic Value of a Statistical Life in Departmental Analyses, Office of the Secretary of Transportation, 2008
- Recommended Interim Values for the Value of Preventing a Fatality in DG Environment Cost Benefit Analysis, European Commission, 2001
- 2005 Valuation of the Benefits of Prevention of Road Accidents and Casualties, Department for Transport, January 2007
- Measuring Effectiveness and Cost Effectiveness: the QALY, National Institute for Health and Clinical Excellence
- According the House of Commons Health Committee the thresholds are not evidence-based
- David M. Cutler, Allison B. Rosen, and Sandeep Vijan: The Value of Medical Spending in the United States, 1960–2000, The New England Journal of Medecine, August 31 2006
- Gordon H. Guyatt, David H. Feeny and Donald L. Patrick: Measuring Health-Related Quality of Life, Annals of Internal Medicine, April 1993
- Stuart R. Walker and Rachel Rosser (eds): Quality of Life Assessment: Key Issues in the 1990s, Springer, 1993  (Google Books extract)
- World Health Organization Quality of Life (WHOQoL) Field Centre, University of Melbourne, January 2007
- EQ-5D™: A Standardised instrument for use as a Measure of Health Outcome, (Euroqal website)
- The Benefits and Costs of the Clean Air Act 1990 to 2010: EPA Report to Congress, United States Environmental Protection Agency
- Economic analysis to inform the Air Quality Strategy, UK Department of the Environment, 2007
- Health Impacts of Transport Emissions in Australia: Economic Costs, Australian Government, 2005
- Bin Jalaludin1, Glenn Salkeld, Geoff Morgan4, Tom Beer, and Yasir Bin Nisar: A Methodology for Cost-Benefit Analysis of Ambient Air Pollution Health Impacts, Commonwealth of Australia 2008 (includes health costs of pollutants from 8 US, UK and Australian studies on tables 5 -12 on pages 99,117,129,138,161,182,192 and 204)
- M. Delucchi and S. Hsu: The External Damage Cost of Direct Noise from Motor Vehicles, Institute of Transportation Studies, University of California, February 1998
- Noise,Transportation Cost and Benefit Analysis II – Noise Costs Victoria Transport Policy Institute
- Assessing an IPCC assessment. An analysis of statements on projected regional impacts in the 2007 report', paragraph 3.3, Netherlands Environmental Assessment Agency, 2010
- Stern Review on the Economics of Climate Change - Final Report HM Treasury 2006
- see the paragraph on decision making in the article on politics