Compounding rewards as an argument for moral virtue

2021-12-26

Much has been written about how actions tend to compound. While a decision is often judged by the immediately visible price, the total cost paid can be far larger, because any particular decision can reinforce all sorts of feedback loops and systems. This compounding nature of actions means that the ‘price’ (often something like expending effort, getting money, losing job stability) is only one part of the total cost. And in some actions that compound unusually well, it ends up being a vanishingly small part.

Take drinking as an example. Not only does drinking alcohol give you a hangover, but crucially, it might also reduce the quality of your friend group, slightly decrease your health, or make you more likely to drink in the future. These effects have yet more effects, and it could go on for arbitrarily many steps. The first order consequence was small, yet the full summation of the consequences was unexpectedly big.

You can analyse any single action like this. Hold up a decision to do something and look into its first order consequences, then the second, and the third, and even then you might only have enumerated only some small part of the impact that the decision will carry. Not all decisions are like this, and most may just fizzle out quickly, but it’s not obvious at all how to predict which is which.

This tendency of actions to compound, and the challenge of predicting which ones will, might be one explanation for a particular flavour of morality obsession that can be seen in great people of the past.

In his autobiography, John Stuart Mill claims that, in his education, moral influences were “so much more important than all the others”. So much so, that he dedicates an entire chapter to his long journey of moral development and its influences. One particular example was how his father strictly guarded him against “self conceit” with “extreme vigilance”, keeping him out of the way of various evils like hearing himself being praised or making “self-flattering comparisons”.

Another example is Benjamin Franklin, whose autobiography has a chapter literally titled “Plan for Attaining Moral Perfection”. It starts like so:

“It was about this time I conceived the bold and arduous project of arriving at moral perfection. I wished to live without committing any fault at any time; I would conquer all that either natural inclination, custom, or company might lead me into. As I knew, or thought I knew, what was right and wrong, I did not see why I might not always do the one and avoid the other.”

He lists thirteen specific virtues that he derived from his readings, such as Industry, to “lose no time”, or Silence, to “avoid trifling conversation”. And this is the best part - he even decides to track his progress on a weekly grid, and sincerely tries to optimise his moral virtuousness each week.

I find something the sincerity of this remarkable. While talk of morality often seems like abstract philosophising, Benjamin Franklin is no mere talker. He actually earnestly optimises for the fulfilment of his virtues! He even does it on a grid! And the implication is that he becomes inordinately successful in no small part due to his character and moral upstanding.

This kind of thinking seems underrated. One reason why is that there is a general lack of respect for decisions that look small. In this frame, every decision should be thought less in terms of its immediate side effects and more as a representation of all the consequences that can come after. This makes even the most unimportant decisions gain a sense of urgency and importance, and makes bad decisions seem worse. But that’s because they are actually that bad: there may be bad decisions that you still haven’t stopped paying off.

It’s true that being obsessed with morality seems to be something of a fashion of their times. But being sincere and earnest about moral virtues might have swung back in the pendulum to become actually underrated in our age. Much of our society now features sophisticated-seeming cynicism about declaring anything a virtue, or even trying very hard to figure it out. While I hold no opinion about whether virtue ethics is useful for achieving outcomes for groups of people, or the process to discover the correct virtues (though the Benjamin Franklin style approach does sound good), I am saying that they can be instrumentally useful to achieve your ends.

For one, moral virtue does have the property of giving you very clear solutions to a large class of problems. If you can’t drink alcohol, you can’t drink, not even a little bit. This circumvents many problems of careful moderation. That’s a good thing: in a world of reinforcing feedback loops, moderation should be treated with a little suspicion.

Secondly, what many people call “luck” can generally be traced back to the compounding effect of a long series of correctly made decisions. Each decision seems small and insignificant at first, but the resulting consequences are large. Moral virtue has the advantage of taking every decision seriously and so could be an unusually effective model to apply when thinking about a world where actions multiply themselves and compound.

Thanks to Sam Enright for reviewing a draft of this, and Neil Shevlin for some comments.