Trying to Make Sense of Kahneman

I have been thinking about the ideas of Daniel Kahneman. I think what he is doing is revolutionary for ethics and political philosophy, and I keep trying to say what his experiments mean. Here is one preliminary stab at sorting that out.

1. A Random Walk Down Wall Street says that no method consistently predicts increases in stock prices in a way that beats very simple strategies or even random selection.

2. Kahneman’s work says that people are not very good at predicting their own happiness. It also says that it will be problematic, in any given case, to say whether someone has in fact achieved a particular level of happiness, since at least two quite different measures of happiness have intuitive appeal. Perhaps the way of putting these two ideas together is that, apart from some extreme and simple maxims (like “avoid severe depression and chronic pain”), it will always be extremely controversial whether any plausible strategy or maxim is superior to any other in bringing about happiness – in oneself or any other rational being.

3. Any personal policy aiming at maximizing one’s own happiness and any moral policy aiming at maximizing the happiness of other rational beings is, for Kahneman’s reasons, extremely dubious, when it goes beyond the obvious and universally accepted points.

4. So, if anything besides the happiness of rational beings has moral weight, that consideration has, in many cases, decisive moral weight and decisive practical weight. That is, if I can’t estimate or predict my own happiness, it doesn’t make sense, beyond making some simple and obvious provisions, to concern myself with it. It is rational for even someone who is committed to always putting him or her self first to allow non-self-regarding considerations to determine action most of the time, as long as that person’s value system gives those considerations any weight at all.

5. Similarly, if one gives any weight to the suffering of non-rational beings like cows and chickens, then, again, that consideration will often be decisive, both for the person whose basic commitments are selfish and for the person whose basic commitments are utilitarian. If I can’t predict my own happiness or that of any other rational person well enough to plot a satisfactory intervention to improve that happiness overall and in the long run, then the clear unhappiness of even a greatly discounted being is decisive.

6. Maybe most important: if I am not a utilitarian of the Mill persuasion, if I do not take the welfare of sentient beings as the only value in the universe, then all sorts of other values may be decisive, in those cases in which the happiness of rational beings is ruled indeterminate. For example, the value of compassion, the aesthetic revulsion against certain sorts of animal raising practices, may carry the argument, once other considerations are bracketed.

7. There’s an analogy in physics. Physics recognizes a number of forces. Some are very strong and work at very small distances. Others are weak but work at very large distances. Suppose that human welfare trumps everything else, morally, whenever it can be clearly evaluated. Suppose that my own welfare trumps everything else, psychologically, whenever it can be clearly evaluated. That might still leave an enormous range of play for other moral considerations.

Here is another effort, in the same direction: 

Is good advice about investment also good advice about life?

Malkiel’s A Random Walk Down Wall Street makes the point that the most popular and plausible strategies for predicting long term market outcomes don’t produce results as good as a random portfolio or as simply owning a fund that buys some index of the whole market – so that one is owning all the stocks and profiting or losing as the market as a whole profits or loses. Another important point he makes: trying to shift one’s investments to get short-term advantages generally costs more in transaction costs than it earns. The only reasonable investment strategy is to hold stocks for a long time.

Kahneman’s work on happiness suggests that our ability to predict the effect of any event or action on our long-term happiness is limited in something like the way our ability to predict the growth in value of stocks is limited. The reasons for the limits are different: we have couple of standards of happiness that diverge, and so the question, “How happy has my life been?” often gets divergent, plausible answers. Also, human beings make systematic mistakes in predicting how they will answer this question. Taken together, these two considerations may make the prediction of happiness as uncertain as the prediction of the long term fate of stocks. What I am interested in: to see what reasonable strategy looks like, if life planning is as uncertain as stock prediction. Would any of the strategic advice that stock analysts give for reasonable behavior given the proven human incapacity to predict the market carry over to life planning?

Whatever one’s goals, the problem of living is somewhat like an investment problem. One takes on projects, hoping one won’t regret having taken them on, knowing that sincere people have come to regret taking on particular projects.

What pieces of advice from Malkiel’s A Random Walk Down Wall Street might have life-planning significance:

Buy and hold. This is standard moral advice and also standard investment advice: don’t be deterred from your projects by setbacks. Don’t jump around frivolously among projects.

Avoid transaction costs. A very conservative person might be just very conscious of transaction costs. (My dad was very conscious of transaction costs.) I think of the costs of some big transitions: divorce, job change, relocation, sex reassignment, upsizing or downsizing one’s house, buying a cabin. Different transactions have different costs, and sometimes one can make quite a bit of change at quite a small cost: starting an exercise program.

It is important to think about the coin of the cost. Usually anything that costs money nicks projects by stealing resources from them. Sometimes transactions cost creative energy or leisure or personal capital or relationships. Sometimes what looks like a cost may actually be a benefit with respect to what the person values, as a Buddhist practitioner might find problems and annoyance to be opportunities to tame the ego, or an adventurous person might take difficulties to be just further adventures --- or a writer might find difficulties to be just good material for the upcoming memoir. (It is hard to think  of how transaction costs work for writers; they seem to be able to turn lots of problems to their advantage. That might be an argument for being a writer.)

Buy an index fund of the market. This seems to be the hard one: there is no moral equivalent to buying an index of the market, to investing in every position: the Nazis on Monday, the resistance on Tuesday, the apolitical hedonists on Wednesday. What is reasonable for an investor is impossible for a sane human being. To own an index comes close to what Plato describes as the situation of the democratic soul. The consequence of this situation: the soul is torn apart, and the soul eventually opts for tyranny – the utter dominance of one interest or impulse.

The idea of a random walk. This may be the idea that takes over the moral discussion, when buying index funds turns out to have no moral analogue. Ethical choices, like choices made by flipping a coin, may be  random with respect to theories about outcome. Maybe that means that one can choose in moral ways, even if one is the sort of person who is primarily concerned to be as happy as possible, who will never do anything long-term to thwart the project of being as happy as possible. If the strategy “choose by moral conviction” is more likely to result in happiness than any strategy that seeks to maximize happiness, then one has no reason, as a happiness maximizer, not to be moral. And the justification for being moral being more likely to result in happiness is not some asserted connection between morality and happiness but simply the assertion that it is random with respect to any theory about promoting happiness – the equivalent of a monkey throwing darts at the stock list or an investor buying an index of the whole market.

A diversified portfolio. This has some of the same problems as the index of the market. What is the counterpart in life to investing in bonds – or in stocks on foreign exchanges? Surely, some attention to non-Western traditions makes sense under this head. One doesn’t want to be stuck with a frame of reference that might turn up inadequate to the next challenge or turn up bankrupt. One can maintain a friendly interest in some other tradition, while holding fast to one’s own – keeping the other tradition in reserve, noting its strengths, probing one’s difficulties with it.

It isn’t crazy to take as one’s moral guideline: have a variety of projects, and have projects of different sorts, within one’s overall value scheme. Someone concerned for justice might work for a political candidate, volunteer at a food shelf, write letters to the paper, start a consumer petition, arrange for his son or daughter to work in Nicaragua for a summer, even write a novel. That’s all pretty normal stuff.

Another thing: commitments are sustained by passion. To keep working on model trains or world peace in the spirit of “carrying forward a previous commitment” (“holding one’s stocks”) might be: to not work hard enough to count as working. With stocks, you either own them or you don’t. 

One might say some things about the relationship between gardening and investment. In some ways, they are close: one plants things with different requirements, not too much of any one thing, realizing that one cannot control the weather. There are several optimistic scenarios for the garden. But one cannot usually pull up one crop and plant another, in response to changing conditions. One necessarily invests for the season.

A corollary: is there an argument for being a moral investor contained in the observation that people who try to maximize their profits don’t do as well as the market taken as a whole – or as a random basket of stocks chosen by a dart throwing monkey would do.  Maybe: if selecting a large group of stocks on the basis, “I like the ethical smell of these” approaches a random selection, then that formula will do as well as any random selection.

Suppose one did the research. Suppose one had people choose a large number of stocks on purely moral grounds, throwing out the companies that seemed to do harm, selecting the companies that seemed to do good. One would, of course, get different baskets of stocks depending on people’s research time and research sophistication, so there would be lots of different moral baskets. One could just look and see how those baskets did, compared to the market as a whole. If it turned out that those baskets did worse over the long haul, one might mix in standard index funds until one reached a mix one was comfortable with, a mix that represented the price one was willing to pay for being ethical. Perhaps I am willing to pay 1%.  So, if my best long-term strategy is to buy an index fund and hold it, I might buy a combination of an ethical fund and an index fund in proportions such that my return was 1% below what the index fund alone would give me.

That feels like an ethically mediocre position. But that’s tricky. This is a policy one can recommend, one that lots of people could adopt, whereas a purist policy like “Only invest in companies you approve of” is not a policy lots of people are psychologically capable of adopting. Even if one is, oneself, able to adopt such a policy, one is probably serving the aggregate of good causes better by modeling and recommending the less pure policy.

This is a bit like the fight in the animal rights world about veganism  campaigns versus “cage-free eggs” campaigns. One has considerable hope of converting large numbers of  people to cage-free eggs, and thereby changing institutional behavior; one has close to zero hope of converting large numbers of people to vegan principles.