Over abstraction redux

From a colleague's e-mail:

George,
In your blog about over abstraction you made a big deal about consistency, at least in the realm of math. The whole point, I thought, was to warn against abstractions that were more wishful thinking than real. But your last blog on your abstract economy model, isn't that in danger of over abstraction? In other words, are you failing your own consistency test?

Let's see if he is right.

My colleague is more of a mathematician than am I, so he naturally focused on that part of the blog. But my real point was more about how over abstraction leads to an inability to ground an abstraction (eventually) in reality. I claim that my abstract economy model is grounded in that it starts with physical principles regarding work and energy flows. That part of the model is only abstract in that I did not attempt to parameterize the flows with any real time scales since there is much we don't know about actual peaking of the fossil fuels. As far as the depletion dynamics and the increasing difficulty encountered in obtaining the next unit of gross energy, that part is real and has been demonstrated for oil in several countries, including the United States.

The real abstraction starts with the idea of assets and their production. In a sense my colleague is correct. I have substituted a wholly abstract relation for the real one and there is no guarantee that I did it correctly. However, and this is the beauty of science, the abstraction is testable in the real world. We can collect data and see if the relation holds in the real world. In fact, one of Charlie Hall's grad students is doing just that. He is going to look at global steel and concrete production as surrogates for the broader base of assets and compare them with net energy production based on declining EROI for the last fifty years. Our hypothesis is that during the last twenty years, or so, the gap between net energy and steel/concrete accumulation (measured in emergy units) has narrowed or actually crossed over as per the model dynamics. If so, we will have made a beginning case for the reality pointed to by the model. It will not have been over abstracted. If not, then we go back to the drawing board. What we don't do is build an elaborate theory of economics out of a simple model without ever knowing if it corresponds to reality.

Of course, some will argue that economists followed the same path. They are forever testing their hypothesis by collecting data and plotting it to see if their projections match the regressions they come up with. And that is partially true. But it points to a subtle error in doing science that I hope we can avoid. It will be hard because scientists can do it unconsciously. And that is picking data that confirm the hypothesis a priori. It takes a lot of caution, skepticism, and eye balls to avoid the pitfalls of such errors. My main take on practitioners within many mainstream schools of economics is that they so thoroughly believed their school's dogma that they were more like religious zealots than scientists in pursuing the truth. Once again, witness Alan Greespan's confession before Congress that he was in shock that markets didn't work the way he had always believed. [Note: such zealotry pertains particularly to Randian market fundamentalists who are so convinced of the ability of free markets that they never question anything about the premise. The parallel with having a religious belief is stronger still. Since there never has been anything approximating a truly free market, just as there has never been convincing evidence of knowledge of God's will, there is nothing to refute or disprove.]

Time will tell if my abstract model of the economy is too abstract. One hopes we will know sooner than later.

What this has gotten me to thinking about, however, is a further explication of some other abstractions that can be used to build more detailed models of the economy. I've already written about the abstraction called money, for example, and this is along the lines I have in mind. I want to further explore debt as a set of layered abstractions on top of money. The simplest statement of this is that debt represents a promise to pay back a monetized value at some specified point in the future. In my next blog, assuming real world events don't intervene, I will start looking at debt and jobs/income as abstract models in the hopes that we can see a way to better understand our current predicaments in the financial and jobs markets.

Comments

[Dave]

What exactly is the "real world"?

[George Mobus]

Dave,

If this is a serious question you will need to provide a little more background on what motivates it. I think most people have a good sense of what "real" means in this context. If you have some deeper philosophical concern then please do share it.

George

[Dave]

I believe in this context the "real world" is the falsifiable, scientific world based on physical reality; as opposed to a world construct based on language, concepts, and metaphysics such as economics or religion. It's just me, but I sometimes find that important to point out.

[George Mobus]

Dave,

Got it and agree.

George