SOPhiA 2019

Salzburgiense Concilium Omnibus Philosophis Analyticis

SOPhiA ToolsDE-pageEN-page

Programme - Talk

Distributional Weights in Cost-Benefit Analysis
(Plenary Lecture, English, Location: HS E.002)

Where a possible government project would involve both benefits and costs, cost-benefit analysis seeks to determine whether or not the benefits are worth the costs, and thus whether the project should go ahead. Usually, the project would on balance benefit some people, and would on balance harm other people; it is therefore relevant in particular how benefits/costs are aggregated across people.
Theorists disagree about whether this aggregation should be "unweighted", i.e. should simply sum individuals' willingness to pay (in money terms) to have the project go ahead, or whether instead one should use "distributional weights", so that a benefit or cost of a fixed size (in money terms) counts for more when it accrues to a poorer than to a richer person. Put differently (but more roughly), defenders of the unweighted approach (resp. the weighted approach) advocate measuring benefits and costs in dollar terms (resp. in welfare terms), for the purpose of straightforwardly summing across individuals.
Among moral philosophers in particular, there is an overwhelming consensus that the weighted approach is the correct one, where the weights are straightforwardly derived from whichever social welfare function one seeks to maximise. For example, if one adopts a utilitarian social welfare function, it seems to follow that the weights (for CBA purposes) should simply be the marginal utilities of consumption, at the various applicable baseline consumption levels.
I will discuss sources of resistance to this consensus, stemming from the observation that various actors (both governmental and private) will react in predictable ways to the government's decisions as to which projects to fund. In particular, it can happen that a project that naively passes a weighted CBA test (and so naively seems to increase social welfare) actually reduces social welfare, once these responses are taken into account.

Chair: Christian Feldbacher-Escamilla
Time: 10:15-11:45, 18 September 2019 (Wednesday)
Location: HS E.002

Hilary Greaves 
(University of Oxford, UK)



Testability and Meaning deco