The Evidentialist's Wager

Suppose that an altruistic agent who is uncertain between evidential and causal decision theory finds herself in a situation where these theories give conflicting verdicts. We argue that even if she has significantly higher credence in CDT, she should nevertheless act in accordance with EDT. First, we claim that the appropriate response to normative uncertainty is to hedge one's bets. That is, if the stakes are much higher on one theory than another, and the credences you assign to each of these theories are not very different, then it is appropriate to choose the option that performs best on the high-stakes theory. Second, we show that, given the assumption of altruism, the existence of correlated decision makers will increase the stakes for EDT but leave the stakes for CDT unaffected. Together these two claims imply that whenever there are sufficiently many correlated agents, the appropriate response is to act in accordance with EDT.

An argument against causal decision theory

This paper develops an argument against causal decision theory. I formulate a principle of preference, which I call the Guaranteed Principle. I argue that the preferences of rational agents satisfy the Guaranteed Principle, that the preferences of agents who embody causal decision theory do not, and hence that causal decision theory is false.

Newcomb dilemma in development management

Newcomb dilemmas show a discrepancy in our rational reasoning, as made clear by comparing Evidential Decision Theory with Causal Decision Theory. In this paper, I look at three versions of the dilemma: the original, highly technical and abstract one plus two more mundane cases. I also account for the general schema of the dilemma possibly appearing in macroeconomic situations. Ahmed (2014) aims to provide a solution for macroeconomic cases that opens room for forming a development management Newcomb dilemma – an imaginary case of electric motor competition between Toyota and Tesla. I argue that Ahmed’s solution may solve the macroeconomic Newcomb dilemma, but it cannot be applied to the development management dilemma. If I am right, similar Newcomb situations could be cropping up regularly in development management, leading to seemingly insoluble strategic decisions having to be made. This may create an inevitable pitfall for development management.

Agency, Intervention, and the Past

Although agency theories of causation are unsuccessful, they draw on two plausible contributions to the analysis of causation: a characterization of agent non-symmetry in terms of effective means and an insight into nature of the similarity weighting for counterfactuals. A development of evidential decision theory provides the most immediately plausible way of understanding agency asymmetry. However, a problem with the proposed development reveals the importance of causal thinking—captured in causal decision theory—in characterizing when an action fails to be the most effective means to a certain end. Non-reductive interventionist approaches to causation are unnecessary because the recommended similarity weighting captures the appropriate notion of intervention. The recommended approach to causal non-symmetry can explain the fact that, metaphysically necessarily, causes usually precede their effect because temporal direction is preponderant causal direction. The non-symmetry of agency is related to two de facto asymmetries relating to knowledge and intervention.

Cheating Death in Damascus

Evidential Decision Theory (EDT) and Causal Decision Theory (CDT) are the leading contenders as theories of rational action, but both face counterexamples. We present some new counterexamples, including one in which the optimal action is causally dominated. We also present a novel decision theory, Functional Decision Theory (FDT), which simultaneously solves both sets of counterexamples. Instead of considering which physical action of theirs would give rise to the best outcomes, FDT agents consider which output of their decision function would give rise to the best outcome. This theory relies on a notion of subjunctive dependence, where multiple implementations of the same mathematical function are considered (even counterfactually) to have identical results for logical rather than causal reasons. Taking these subjunctive dependencies into account allows FDT agents to outperform CDT and EDT agents in, for example, the presence of accurate predictors.