Abstract: According to the standard view, the temporal direction of causation consists in a quantitative asymmetry: Causation is widespread in the forward direction but (almost) entirely absent in the backward direction. Earlier events frequently cause later events but later events (almost) never cause earlier events. In this paper, I argue against the standard view. The direction of causation does not consists in a quantitative asymmetry.
    The causal facts supervene on the physical facts. So if causation is quantitatively asymmetric, the physical facts have to contain a suitable temporal asymmetry on which the quantitative asymmetry of causation supervenes. I argue that the physical facts do not contain any temporal asymmetry that is a plausible supervenience base for a quantitative asymmetry of causation. Hence, the intuitive direction of causation cannot consists in a quantitative asymmetry.
    I sketch an alternative metaphysics of the causal direction according to which it consists in a qualitative asymmetry. Causation is widespread in both temporal directions but has a different character in the forward than in the backward direction. The qualitative difference explains why causation is, for creatures with capabilities and goals like ours', only of practical interest in the forward direction.