The pursuit of happiness as a forecasting problem

So many have written so much for so long about the pursuit of happiness, without, it seems, having solved the problem definitively. This suggests one of (at least) two things: perhaps the problem has been posed so poorly that no solution makes sense (Q: “What is the meaning of life?” A: “42!”) or perhaps there are many valid solutions and our quest for the solution is the problem. I’ll make a case for the second option, by formulating the pursuit of happiness as a forecasting problem.

We can think of happiness at any point in time as an increasing function of how well realized outcomes correspond to (or exceed) desires at that time. One path to happiness therefore is to have no desires, and that is exactly what some schools of thought seem to recommend.

However, another approach begins by asking what exactly are you doing when you are pursuing happiness? I think it is fair to say you are taking actions that you hope will generate outcomes that will satisfy or exceed your desires. But at which time?

A hedonist would want to maximize happiness for as long as possible into the future (suitably discounted). A diarist on the other hand might want to maximize peak happiness levels, to be recalled fondly later (since our memories seem to be shaped more by peaks and end points than the average over time). A legacy seeker might aim to maximize a happiness she will no longer be alive to experience. One could be all three and perhaps more, of course, in some mixture.

But what matters for my argument is that in each case, the pursuit of happiness requires a forecast about outcomes and desires at some future point(s) in time. This is a hard problem, for at least two reasons.

Causal complexity

First, the causal links between our actions and the outcomes that we are aiming for can be very hard for us to comprehend. Time lags between the actions we take and the outcomes we are hoping for are typical, and there will be often other factors at work that affect the outcomes that we have no control over or even awareness of. So causal complexity — which makes it hard to know whether a course of actions will yield a particular outcome- is the first big forecasting challenge to the pursuit of happiness.

Preference plasticity

Second, even if we could accurately forecast which actions to take to produce a desired outcome, we might discover that on the day the outcome arrives, it is no longer what is desired. Our preferences are not fixed in stone (even if they get somewhat harder to change as we age), and it’s hard to tell what our preferences will be at some future point.

Now, it is true that we will become aware over time whether our actions are bringing us closer or not to some objective, and our preferences might adjust to stay in alignment (this powerful psychological force is called “dissonance reduction”). However, if this adjustment process worked perfectly- we could always learn to desire the outcomes that we realized- we would be surrounded by a sea of happy people, but we are not. Further, there are also psychological forces that can cause our desires to change rapidly once they are close to being satisfied (these are known by labels such as “satiation” and the “hedonic treadmill”). For my argument, the details of how this happens do not matter that much; what does is that preference plasticity creates a second and significant forecasting challenge to the pursuit of happiness.

So, when is the pursuit of happiness likely to be successful?

If this account is correct, then all else being equal, those who find it easier to forecast the outcomes that will arise from their actions and their future preferences are more likely to be happy.

I am ignoring the (rare?) possibility that errors in predicting outcomes and desires cancel out, producing happiness by accident. I am also ignoring the case of happiness that arises “for free” when desirable outcomes materialize without any effort of our own, because my focus is on the pursuit of happiness and the conditions for its success. That, I propose, is equivalent to success at forecasting.

Does this mean that being smarter makes it likely you will be happier? Not necessarily- because we are not all solving forecasting problems of the same level of difficulty.

For instance, an individual with low preference plasticity should face an easier forecasting problem. They can take for granted their future desires and need to focus only on forecasting what actions to take to meet them. Perhaps growing up in a cultural context with strong norms (or just becoming older) reduces preference plasticity, and so makes for more successful pursuit of happiness. There might also be a genetic component to variation in this property across people.

Similarly, objectives with shorter times lags and more controllable causal links (between actions and objectives) should also lead to an easier forecasting problem. Life in systems that do not see much innovation or change, where causal links remains stable and well understood across generations, may therefore be more conducive to the successful pursuit of happiness. Finding goals that one can pursue with a degree of competence (which might be harder or easier for others) is also a path to happiness. Focusing on the accomplishment of immediate tasks at hand and “enjoying the journey rather than hankering for the destination” seem to be versions of this idea.

Viewing the pursuit of happiness as a forecasting problem explains why there seem to be so many different ways in which people might be happy (or unhappy). It also does not commit us to any particular value system or ideology — I have not said anything here about the nature of the outcomes and preferences — whether they are selfish or altruistic, whether they are good or bad for the world, whether they involve finding meaning and connection or instead, nihilism and isolation. I do not doubt these things matter too for a group living species like ours, but I wanted to focus here purely on the forecasting problem and why its typically hard to solve.

A closing conundrum: what about those individuals who develop a meta-awareness about the pursuit of happiness being a forecasting problem- i.e., they can correctly forecast that it will be hard to forecast outcomes and desires? Are they doomed to unhappiness through anxiety, or do they transcend the pursuit of happiness altogether, recognizing its essential futility?

The measurement of happiness has become an active area of social science research. A nice compilation of measures can be found here: https://www.pursuit-of-happiness.org/science-of-happiness/measuring-happiness/. Writings on happiness have proliferated as well, and I am grateful to TH for sending me a particularly annoying specimen, which provoked my own attempt at contributing to this genre.