Any social-contract theory bases justice on reciprocal interaction. Justice is what rational self-interested people would agree to under fair conditi ons. As we cannot interact with people in the far-distant future, a social contract with them seems impossible. We hold their quality of life, and their very existence, in our hands, while future people can offer us nothing in return.

The standard affluent test case was the time bomb: an action that is beneficial to present people, devastating for distant-future people and irrelevant to intervening generations. Parfit's risky policy was a classic example. Our choice of the cheaper power plant is good for us, has no effect on anyone for several centuries and is then very bad for those who are alive when the radiati on leaks. Surely planting a time bomb is wrong, especially if the present benefit is negligible. (Planting a time bomb purely on a whim would be morally equivalent to driving a carbon-fuelled vehicle just for fun!) But how can any contract or agreement between self-interested people condemn it?

Some affluent social-contract theorists bit the bullet here. If justice requires interaction, then we have no obligati ons of justice to future people. We may happen to care about them, we may choose to take their interests into account, but we owe them nothing.