On Ethics - Part 1 - Utilitarianism
Does Rule Utilitarianism provide a plausible alternative to Act Utilitarianism?
Yes, no and maybe.
It can be a valid alternative under some conditions:
- The comparison is drawn over maximising happiness of a social group.
- The two are compared across time – with many actions/consequences for evaluation.
- The rules are followed, only with defined exceptions.
On the contrary, Rule Utilitarianism would be an inadequate substitute to Act Utilitarianism – in maximising happiness – if the inverse is true.
Both: Act Utilitarianism (AU) and Rule Utilitarianism (RU) try to help us realise happiness. They both hold that it is life’s only desirable end. But they crucially differ in that Act Utilitarianism judges individual acts by how well they promote happiness, where as Rule Utilitarianism judges rules, that dictate actions, by how well they tend to promote happiness.
Like with any moral framework, AU and RU are both useful fictions; in that they matter only in-so-far that they help humans satisfy things that we value; wether thats pleasure, justice, rights, etc.
It is in the success of realising what we value, then (namely happiness), that RU and AU become comparable.
When asking wether RU can be a plausible alternative to AU, we are really asking wether one can better optimise for happiness than the other, since this is their only shared criterion. Analysing the mechanism of each moral framework more rigorously; RU is best understood to maximise expected utility across the distribution of actions rather than that of any single case.Whilst AU maximises expected utility within each particular case rather than across a wider distribution of actions.
RU was introduced to fix the AU’s dilema, that individual actions that try to promote happyness, can demote it in large social settings. A Peeping Tom being an illustrative example. A peeper may well promote his happiness through the voyeuristic act, but in reality, if we suppose everyone knows that other agents might reason this way, it could still produce a society with more anxiety, distrust, and unhappiness overall. Rules like “don’t peep, unless it’s necessary to prevent serious harm”, if gone unbroken, solve this dilema, and can better optimize for long-term social happiness.
It can objected that AU does not necessarily ignore long-term or society-wide effects. An act utilitarian evaluating whether to peep can count anxiety, weakened trust, the probability of discovery, effects on social norms, precedent, and the consequences of others adopting similar reasoning. This is true. A sufficiently sophisticated AU oracle could in principle calculate all of these wider consequences and arrive at the same prohibition as RU. But this weakens AU as a practical moral framework for ordinary agents, as in the real world, people struggle to reliably calculate consequences of their actions. On the contrary, this is precisely where RU shines, by avoiding the epistemic burden on the agents, by providing rules that can generalise across classes of actions.
Perhaps a stronger objection is that even an excellent rule can cause catastrophic consequences in an edge-case. More perversely, if one know the rule might cause a catastrophe, must they still blindly follow the it? That indeed seems irrational even by RU standards.
To that I’d counter-object that rules can rarely define general exceptions within them, that would prevent such twisted cases. “Don’t lie unless it will prevent serious harm” being a classic example.
The substance is in whether adopting and generally following the qualified rule produces more happiness overall. I think the difficulty is mainly in how such exceptions should be defined and how specific they need to be.
The verdict then, is that for an omnipotent calculator, AU is as good as RU. With perfect information, act utilitarianism already counts every distant ripple.
But for biased, imperfect, social beings, generalised rules with coarse exceptions, if judged across many cases, are a suitable alternative.