An Economy Stalled?

No matter how much the Fed injects money into the system, it doesn’t seem to be responding. The U.S. economy right now feels like an airplane that has just stalled: it’s still pointed up, but instead of going up its altitude is decreasing fast, tail first. All the trillions of dollars of stimulus is not providing enough thrust to push this airplane up, and the momentum so far is still downward. That false lift we felt the past weeks in the stock market may be just the start of a flip.

When an airplane stalls, it’s difficult to put it back in control. It requires a lot of skill on the part of the pilot to put it back in control. Unfortunately, I think our leaders don’t have the requisite skills and knowledge to right this airplane. Nobody really knows what’s going to happen next. All the macroeconomic models so far devised are useless. Could it be that all of current macroeconomic thinking and its resulting models, even those that consider dynamics (feedback loops), is wrong?

Macroeconomic Models

Macroeconomic models are a necessary tool for economic policy making. The Fed is guided by models; however, these models most likely are not correct. In fact, the whole idea of macroeconomics is suspect.

I propose that macroeconomics is not a science. Like the notorious earth atmospheric models produced by over-zealous environmentalists that predict a non-existent catastrophe, macroeconomic models have so far been useless to policy makers. As proof of this, observe the helplessness of current policy makers in predicting when to stop injecting tons of money into the U.S. economic system. If the amount they inject is not enough, it’s a big problem (their fear in this case is deflation). If, on the other hand, they inject too much and do not stop injecting at the right time, runaway inflation can result. When should they stop? They are watching a number of indicators, and hopefully the indicators will tell them exactly when to stop the money flowing into the economic system.

Just today, White House Economic Adviser Lawrence Summers said as much:

"The thing about an inflation is that … the moment it’s absolutely clear you have the problem is a moment when you may have been too late in addressing it," Summers said. "So I think it’s a very difficult balance the policy is going to have to walk."

Summers claims that, although difficult, it’s a matter of balancing. If you can point to me any economic model that can help policy makers in the current situation and that can tell them exactly what to do, then I stand corrected in stating that macroeconomics is not a science. The fact is that economists steeped in computer modeling themselves will tell you that their models, although very complex, are still of very limited use. Just think: if someone has come up with a good model, he would have started a business with it and made tons of money. A good model is pursued not just by policy makers, but businesses as well.

Aggregate Decisions

Why can’t macroeconomics be properly called a science? Its problem is two-fold:

• There is the philosophical issue of volition. Each human actor in an economy has “free will”, and no mathematical model can capture the statistical significance of this. I cannot predict with much certainty what you will decide to do in a given situation, much as you can’t predict how I would react to the same situation. This is true even in extreme cases like if a gun were pointed at your head and you were being forced to leave your own offspring to die. I cannot predict what you would do, and probably neither can you yourself. We are very complex economic actors.
• There is the issue of aggregate decision and action. Sometimes we behave like lemmings: our individual decisions are influenced by those around us. Given a situation in which we are given very few information by which to decide, we simply follow what everyone else is doing. Any computer model has to “recognize” situations in which the lemming effect can occur, and also recognize those in which it is not possible (as in the case of the secret ballot box). There are other aggregate effects to be considered (follow the leader, group panic, etc).

Only in very limited cases can we assign probabilities of occurrence. One case that comes to mind is nationwide elections. We poll a small percentage of the population, and as long as that small sample is randomly chosen, the predicted results can have a high probability of occurring. But even in this case, multiple feedback loops can render the polling result problematic: the polling result, if published, can affect the final outcome of the election. Potential voters whose candidate is predicted to lose may not vote at all, thereby causing their candidate to lose votes even more. The act of measuring population sentiment can affect that sentiment itself.

When a large majority of economic actors start seeing a recession, most of them will react and reduce their spending. By how much? We can’t predict, even though we can measure after the fact. It is only as a historical fact can we categorically say that “pessimistic outlook has reduced output an additional 10%”.

Feedback Loops

Even our minutest actions are guided by large amounts of information fed back to our senses. The mere act of walking is very complex: every step is a delicate balancing act in which the muscles involved are commanded to give just the right amount of exertion at exactly the right time. Every muscle exertion is a result of very complex calculations with feedback inputs from all the senses. If the next step happens to land on a banana peel, the feedback loop changes its configuration so that the slip signal is given primacy over all others.

Every little decision we make is also the result of a highly complex feedback loop. Therefore, if we take an individual and consider her apart from the rest as an independent economic actor, she is by herself one complex system of feedback loops. As part of an economic system, an individual becomes part of a very complex loop with connections both local and global. If local feedback loops are complex by themselves, imagine the complexity of an economic system in which each actor is also receiving feedback from a large number of global signals. We don’t even understand how our brains work, how can we claim to understand a system with hundreds of millions of brains!