At the Margin

The title of this blog post is a very difficult one for most people to understand. I have been thinking about why that is. And I think I have an answer that I find satisfying. (And yes, after a long hiatus, I'm dropping one of those long, rambling posts.)

Let's start with a simple hypothetical. You commute to work every day by car. There are two routes from your home to work that are fairly similar (parallel roads, say). If you drove each route multiple times, and recorded how long each route took each time, after a while you could probably conclude that one route was 'faster' than the other. But it is also likely that, at the end of the process, if the routes really are similar, that they don't really take differing amounts of time. That is, the average commute time for each would be, with a large enough sample, similar. IOW, it wouldn't really matter which route you took; your commute, you can expect, on any given day, will take so much time. There is little that you can do to speed it up--in the aggregate.

Aha, but we don't experience life in the aggregate, do we? Indeed, you would be a strange person to actually record commute times and try to determine which route is faster than the other. On any given day, myriad factors affect your commute. Some are factors one might include in a multivariate model (e.g., time of day, weather, even season, depending on geography). You would be even stranger to account for those.

But we also know that, on any given day, there are contingencies that affect commute time. So a traffic accident on one route makes the alternate route faster, for example. So on that day, choosing one route makes a big difference. But without knowing when and where such things will occur, one cannot make the choice in the first place.

Instead of thinking in the aggregate, human beings think largely in narrative. And we tend to be the 'hero' of the narratives we tell. So we are hard-wired, more or less, to think that we can 'do things' to achieve our ends--to speed up our commute, in the example. We can take a side street to avoid a congested spot, for example. (Think about driving in a strange city with a long-time resident, and how they often take unusual routes to get places rather than main streets.)

Now, on any given day, a particular route may actually be faster than another. So sometimes we can 'do something' to achieve our purposes. But, again, there are contingencies, and one cannot know when and where they will occur. So some days the alternate route we devise through long experience may actually be faster. But other days, the main surface streets will be faster.

But our internal narrative machine reminds us of that one day when we got stuck at that light for three turns. That really slowed us down. The anecdote, the striking example, sticks with us. Again, we don't think in the aggregate.

And, back to the title of the post. Even if we can devise a 'faster' route, even if we grant that, how much faster is it? Probably not much faster. A few minutes, maybe. IOW, our narrative machine drives us to think of ways we can 'do something' that work, at best, at the margin.

But we don't experience it as 'at the margin.' We experience it as 'success.' And I think that we experience it as success because, when we are frustrated by something, like a commute time, we find any marginal improvement to be emotionally satisfying. Again, we are the heroes of our internal narrative machines, and when we 'solve a problem,' we are most heroic.

Even if, that is, the solution is of marginal value. The commute still takes about the same length of time. If we reduce the time it takes to drive home from work by 2 percent, on some days, that's a victory.

I would go even further and say that many, many people--even smart people--are under the impression that small changes can produce substantial results. And sometimes they can. But not usually. Small changes result in, at best, small results, most of the time.

So tinkering, 'at the margins,' will produce small results, if we're lucky, 'at the margins.' And we haven't even started to talk about the costs of tinkering in the first place. Because once we factor in the costs associated with tinkering, then the small changes we might actually achieve may not net us anything.

Again, I find that few people think in this way. Many people spend a lot of time and energy contemplating what is, essentially, tinkering at the margins. And expecting great results from doing so. This just appears to be hard-wired into us.