Socio-Economic Entropy

Steven Pinker's Writings
I'm not satisfied with either Pinker or the responses to him on this entropy of wealth stuff.

Here's what I want: If this entropy exists, what would it be? Must figure out:
 * Is he talking about something true, real, useful, important?  In what way?  And, if so, can we be clear what it is?
 * Is it accurate to call it "entropy", and use similar equations to make predictions, understand the causality, and engineer outcomes?  Why or why not?
 * What the macro variables are
 * What the micro-variables are

As for his infamous line about "who is to blame", consider another example involving the increase of entropy:  melting ice cream. The entropy of the ice cream could be preserved or reduced by putting it in the freezer (which exports the entropy of the ice cream to the environment as heat). And we can talk about who is to blame for leaving the ice cream out of the freezer.

Now then, it seems that the macro variable he is talking about is simply wealth. And so the micro variables, what are they? All other variables? Or only those variables that, together, are what we call "wealth"? (see my "what are things", and "name calling" pages for a bit of the relation between micro and macro variables) Or (as many of his critics would do) must the most significant micro-variables be actual temperatures? In that case, the temperatures of what? The people's bodies? The temperature of the "wealth"? Perhaps the temperature of the goods and services that count as wealth?

Well, what is "temperature"? It's a macro variable. It describes the motions of particles. So, for wealth, are we describing the motions of the components of wealth? Or, at least, the degree (on a spectrum) of how much these components are one way or another (some "property" they have), such that if they are more one way, the wealth is higher; but if they are more another way, then the wealth is lower? This seems about right, though vague for now. Seems like this "property" would simply be the value of the component. How valuable a micro variable is to the person.

That does seem a bit like "it's wealth all the way down" though, and we want the micro variables to be different from the macro variable, don't we? Anyways, can probably work in some equations that calculate how valuable something is. It does seem like it would be a "matching problem" (matching states to desires), thus a lot like an entropy problem (like maxwell's demon).

Recently found this video, "A better description of entropy". Says that entropy is how evenly spread out energy is, and it has to be unevenly spread out in order to be useful. I think this is a quite helpful explanation, better than Siggy's here while discussing Pinker's statements. This should make the analysis much simpler and productive.

Stream of consciousness-ish:

Of course, we can (maybe) abandon the limited concept of how spread out energy is. Surely the same statistical laws hold regardless of what sort of "Brownian motion" we are considering. Whether atoms or ping-pong balls, the same probabilities apply.

These two features (spread out-ness of something, and the statistical reasons that spreading out occurs) are predictable (thus useful for making predictions) and are perhaps all we really need here. Now of course "poverty" can be modeled similar to sorting all hot molecules into a particular corner of the room. At least in the sense that what people desire (AKA "wealth") is a vast set of specified variables (we want every moment of our life to be a specific way, not all the random other ways it could be, most of which would be bad, such as being too hot or too cold, or going without food for too long, we want good things close to us and bad things moved far away, etc.). And this vast set of variables is unlikely to just by luck all align to the values we desire. It's more statistically likely many of them will be wrong.

And yes if you identify a person as choosing certain variables wrong, that person can be blameworthy, needing to change. If we can turn "wealth" and poverty into the equations of entropy, then of course we can do the same for morality. Indeed morality, like wealth (or, more broadly, desiderata), is about specifying a vast set of variables to be exactly how we desire them to be. Just one is the set of means, the other is the set of ends.

Now, I hate using the word "wealth" when I think that's not the correct word. Wealth is just a subset of what we want (unless we use "wealth" as a metaphor). Really, it should just be "desiderata".

The basis of entropy
Something I don't see enough of is talk about why a law like entropy holds at all. Most people resort to talking about the probability of certain arrangements of micro states. Or the number of possible arrangements of micro states. And appeals to intuition about stuff like shuffling a deck of cards, or mixing a box full of colored balls, "the law of large numbers".

But why are the cards shuffling randomly? Why are the balls in the box not self-organizing? Why doesn't energy just gravitate towards other energy? (and indeed, how does gravity fit into discussions of entropy?)

It seems to me that something akin to Newton's Laws of motion are required for the "law" of increasing entropy to hold. More specifically, maybe it's the symmetry of action and reaction, like in his 3rd law.

This might not seem better than saying "just because", but (if it's correct) at least I think it fills in the blank with a precise positive description. And this kind of thing might be helpful for anyone who wants to talk about analogies like "socio-economic entropy".

What would be the equivalent ultimate basis in the analogy?

How does the quantity behave?
Does it tend to increase, decrease, or stay the same?

Whether talking about wealth or happiness, is it a constant like the energy of the universe? Is civilization itself not a net gain per person, merely a redistribution of wealth or happiness? If you were behind "the veil of ignorance", would the average be the same, and only the statistical variance would be different between the two options?

Some people seem to think so, as if you can only gain if you take from others. I think that model is implausible. We aren't a closed system, we absorb resources from the environment, and things like economies of scale and specialization and all that seem to indicate the opposite. These kinds of things are the only point to civilization, otherwise we'd literally be just as well off living solitary in the wild. And Steven Pinker himself promotes the idea of human progress, doesn't he?

I also can't imagine the quantity having a tendency of getting worse over time, as if we were actually better off back in the stone age. Again, there are people who seem to think this is true ("anarcho-primitivists", and some new-age and communist types who are really really against "materialism" seem to think this).

Unlike conservation of matter and energy, and unlike the increase of entropy, I personally suspect that living systems will exhibit the third option. They will behave precisely the opposite of what you would expect if you naively applied the metaphor of entropy (which always increases) to them. We will have progress rather than decay. The entropy of the entire universe will increase, but surely within the non-isolated systems known as living systems we see accumulation and storage of energy, and it seems plausible that more abstract measures could likewise show this tendency.

See Also:

 * Social Inefficiencies
 * The Matching Problem