ANU Annual Economics & Philosophy Workshop

Friday 22^nd November 2019

Fred Gruen Seminar Room, Research School of Economics, ANU

Note that all teas and lunch will be served nearby in the inner courtyard.

Schedule:

9.30 – 10.45Stéphane Zuber, Paris School of Economics

“A new puzzle in the social evaluation of risk”

Long tea break

11.15 – 12.30Kirsten Mann, ANU Philosophy

“The Relevance View and non-binary choices”

Lunch break

13.30 – 14.45Geir Asheim, Oslo Economics

“Why guests to Hilbert’s Hotel are unwelcome: Strong anonymity and positional dominance”

Short tea break

15.00 – 16.15Patricia Rich, Bayreuth Philosophy

“Epistemic Non-Conformity as a Cooperative Strategy”

Short tea break

16.30—17.30John Broome, Oxford and ANU Philosophy

“Population, separability and discounting”

Drinks at Fellows

--------

Abstracts

Stéphane Zuber, Paris School of Economics

A new puzzle in the social evaluation of risk (joint with Marc Fleurbaey)

We highlight a new paradox the social evaluation of risk that bears on the evaluation of individual well-being rather than social welfare, but has serious implications for social evaluation. The paradox consists in a tension between rationality, respect for individual preferences, and a principle of informational parsimony that excludes individual risk attitudes from the assessment of riskless situations. No social evaluation criteria can satisfy these three principles. This impossibility result has implications for the evaluation of social welfare under risk, especially when the preferences of some individuals are not known. It generalizes existing impossibility results, while relying on very weak principles of social rationality and respect of individual preferences. We explore the possibilities opened by weakening one of our three principles and discuss the advantages and drawbacks of these different routes.

Kirsten Mann, ANU Philosophy

The Relevance View and non-binary choices

The Relevance View, of which Voorhoeve’s Aggregate Relevant Claims is an influential example, is a theory of limited aggregation. It allows the claims of one group of people to aggregate to outweigh a stronger claim in some cases, but rules out aggregation in other cases. The view has intuitive appeal when applied to simple binary choices, but it runs into trouble in non-binary choices. I will argue that the Relevance View entails a particular account of the structure of the moral reasons that obtain in the cases of interest, which explains why extending the view to non-binary choices is difficult. On this account, a claim’s being satisfied under some option may be a reason to choose that option rather than a specific alternative, but not a reason to choose the option relative to the set of all the alternatives. I will propose an extension of the Relevance View to non-binary choices that is consistent with this account of the moral reasons.

Geir Asheim, Oslo Economics

Why guests to Hilbert’s Hotel are unwelcome: Strong anonymity and positional dominance

[Link to full paper] This paper re-examines the incompatibility of Strong anonymity and Strong Pareto. We insist on Strong anonymity as an axiom of impartiality and ask how far the Paretian principle can be extended without contradicting Strong anonymity. We show that Strong anonymity combined with four rather innocent axioms has two consequences: (i) There is sensitivity for a person’s well-being if and only if a co-finite set of people are strictly better than this person, and (ii) although Hilbert’s paradox of the Grand Hotel shows that adding people to an infinite population is feasible, such addition cannot have positive social value.

Patricia Rich, Bayreuth Philosophy

Epistemic Non-Conformity as a Cooperative Strategy

It is common for people to observe that everyone else thinks X, but nonetheless conclude that not-X or to reserve judgment about X. Such epistemic non-conformity presents a puzzle. On the one hand, there are various arguments to the effect that rational individuals should integrate information from others, often leading them to conform. On the other hand, epistemic non-conformity is fairly widespread, often explicitly valued, and plays an important role in improving epistemic performance on the group level. I aim to resolve the puzzle by casting doubt on the claim that non-conformists are individually irrational, once the social context is taken into account. Towards this end, I share the results of agent-based simulations of groups composed of conformist and non-conformist individuals. Two classes of existing models intersect in these simulations: information cascade models, and prisoner's dilemma models. Information cascades are a paradigmatic case in which rational agents are thought to conform, but non-conformist individuals make groups more epistemically successful. Prisoner's dillemmae are situations in which individuals may cooperate or defect; defecting is dominant in the one-shot game, but cooperation can be better when the game is repeated. In my simulations, cooperation and defection correspond to non-conformity and conformity in information cascades. The general conclusion is that non-conformists can perform well, and under some conditions non-conformity can be the only viable strategy.

John Broome, Oxford and ANU Philosophy

Population, separability and discounting

When economists evaluate the consequences of an action, they typically do it in two steps. First they set a value on people's consumption at each date in the future, or more generally on how people's lives go at each date in the future. Having arrived at a value for each date, they aggregate these values across time in order to arrive at an overall value, generally applying a discount rate. This two-step procedure assumes a particular sort of separability of times. It assumes that there is such a thing as the value of the world at each time, having the property that the overall value of the world is an aggregate of the values at each time.
When we are concerned with a long period, during which some people and born and some people die, the assumption of separability has strong implications for population ethics. It is consistent only with a strong sort of total utilitarianism. This ethical theory is built into the standard methods of economic evaluation.