« Education: what works? | Main | Speaking of leaving things undone ... »

Trends don't continue

All good things must come to an end; bad things too. Things seem to stay the same for a long time, then suddenly shift. Punctuated equilibrium. We didn’t notice the inflection point. I mentioned decline of Hispanic immigration and the unexpected reduction in U.S. CO2 emission elsewhere, things that seemed would never end. Now the prison boom is ending. Prisons are closing because there are fewer prisoners.

I think we are surprised when trends end "so fast" because we tend not to notice the trend at all until it is completely manifest in a mature stage. In other words, trends are often peaking and nearing decline by the time we see them.

There is the old and somewhat cliché story about a pond being covered by lily pads. I will repeat for those who have not heard it. People walk by a pond every day w/o noticing many lily pads. Finally, they notice some and then in a few days it seems that the whole pond is full. This seems like magic, but it is simply the result of normal growth, which under some conditions is essentially exponential. Here is the hypothetical. One lily pad covers 1/100. Two cover 2/100 four cover 4/100; 8 cover 8/100. 16 cover 16/100. This is when you might notice them. The next squaring gives us 64/100 and next the whole thing is covered in what seems no time at all. Trends can collapse just as fast. Extending our lily pad story, by the time we notice that half the pond is covered, they are beginning to shade each other out. The decline is started by the time we see the climax.

The Hispanic tsunami ran out of power because of rapidly falling birthrates in Mexico and better job prospects there. We COULD have seen that back in the 1990s. Dropping U.S. emission came from shifts to natural gas and higher prices for oil. We could have seen that back around 2000. The prison population follows the crime rate. Crime rates were growing until the early 1990s; then they fell. With some lag time, fewer criminals mean fewer inmates in prison. We could have seen this coming by around 2005. What is the next trend that will grow and collapse? If I knew things like that, I would be rich. In retrospect they are easy to see; not so easy in prospect.

What I can predict with near certainty is that many of the predictions that we are making today will look silly in five years and that many of you who believe them now will claim that you knew the changes were coming.

I have been reading an interesting book called "Anti-Fragile". It is by the the same guy who wrote "Black Swan" & "Fooled by Randomness." The main idea is that we cannot really assess risk and certainly cannot predict the future, so the best strategy is to create systems robust enough to absorb the shocks we know will come but cannot predict in sufficient detail to address or avoid. I think that makes sense.


Hosting by Yahoo!

Comments

Great blog post John. At least part of the blame needs to fall on politicians and the 24/7 news cycle for over-hyping the problems for ratings and short-term political gain.

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)