Unpredictable
by Joseph Rowlands

One of the key insights in Austrian Economics is that economic consequences are really unpredictable. This isn't to say that the science hasn't made enough progress to make accurate predictions. It isn't even saying that the predictions must be qualitative in nature as there are no fixed mathematical relationships in people's values and choices. It really means that it is impossible to predict the future.

Consider a well known economic law. If the minimum wage is increased above the market rate, unemployment will increase. This is a very simple relationship. Some workers are of marginal value, meaning the value they add to the company is just barely more than they cost. An increase in minimum wage would mean that these marginal workers would now be priced above the value that they add, and so it no longer makes sense to employ them.

If we understand this as a kind of relationship, there's no problem and we can say it is true. But if it is meant to say that inevitably unemployment will increase, it's just not true. There are many reasons for this.

First, there may not be any marginal employees. It's possible that they all bring in far more value than they earn, even if this is unlikely. It could be that the employers were already planning to increase wages, and so the increase in minimum wage did nothing. In general, we don't really know everything about the current situation, which makes predictions hard.

A second kind of problem can happen if people don't act on the new set of incentives. It may not be profitable to keep an employee hired, but an employer may decide to keep him on anyway and see if business gets better before he has to let him go. So just stating that there is a change in incentives does not guarantee that people will change behaviors in any limited period of time.

A third kind of problem is that other factors may be in play that overwhelm this change anyway. Maybe minimum wage increases, but there was a huge expansion of the money supply and wages increase, or companies appear to be flush with profits creating the illusion that the employees are still adding value. Or new technologies may be cutting costs at the same time. In general, this points to the fact that many other factors may be changing at the same time.

A final problem (final for this list of examples, not necessary the last reason for unpredictability) is that people's preferences may change. The change may be small or large. If people change their buying patterns some companies will do better while some do worse. Some companies won't notice that preferences have changed until their product hits the market in a couple of years and they find nobody is interested. A change in preferences can completely change the outcomes.

The latter problem in particular is a cause of unpredictability. There's no way to determine how people will behave in the future. You can make guesses, of course. But there's no way to know if you're right except by waiting and seeing. You can't accurately predict. You might be way off. An increase in the minimum wage law might end up with a decrease in unemployment, due to many other factors affecting the outcome.

It really is unpredictable. It's not just that we don't have the right rules, or we don't have enough information about the current state of the economy. The fact is that preferences can change, new technologies can be discovered, and individuals may learn new information and act on it. Even if you asked people how they plan to act, all of that can change when they learn something new. They can't predict what they don't know but will know, and so they can't predict how their behavior may change over time.

What's interesting about this inherent unpredictability is that it makes certain methodologies inherently flawed. One view of what is proper in science is an empiricist approach. It suggests observation as the key to understanding. And when it comes to causal relationships, it requires the ability to observe the consequences of an action.

This empiricist approach to economics would suggest seeing what happens when minimum wage increased. It would reject any theories about what should or shouldn't happen. Those would be rejected as armchair philosophy, like ancient philosophers suggesting that heavy objects fell faster than lighter objects. The true mark of science, the empiricist would say, is to observe what actually happens.

So they'd do a study of all of the times minimum wage was increased in various places, and they'd find that sometimes unemployment went up, and sometimes it went down. They might conclude that minimum wage has no effect on unemployment, or they might attempt to isolate the variable more. But if they ended up with the conclusion that an increase to minimum wage decreased unemployment, they would accept it and say the data has spoken.

The Austrians take a different approach. They would not accept that a decrease in unemployment was caused by a rise in the minimum wage. For them, it is not a matter of observation. They know beforehand what the relationship is, and if you saw the opposite, they would know that there are other factors in play.

The empiricist approach treats the cause and effect as connected through some unknown means. If you raise minimum wage, who knows what it'll do to unemployment! The only way to find out it is to try and see. The Austrians see the cause and effect as connected through a specific causal mechanism, and logic dictates the possible results. Instead of having to search for a correlation and assume a causation to go with it, we already know what the causal mechanism is. We can logically deduce it from earlier premises.

Because of the lack of predictability, these two approach are incompatible. The empiricist approach would not be able to determine if there is a casual relationship, or at best would see it as a kind of probability. They would see the deductive approach as being unverifiable, which they would conclude is proof that it is not science.

Of course, it isn't unverifiable. But the process of verifying it is not simply measure the result. The process of verification would be in determining of the logical deductions were valid, and whether the initial premises were true. It certainly can be falsified, but not through a naive approach of measuring outcomes and ignoring the unpredictability of the factors that shape the outcome.

The Austrian approach isn't perfect either. They pick the other side of the false dichotomy between empiricism and rationalism. They believe their premises are axiomatic and a priori. But we don't need to follow them that far. We can accept the initial premises as well-supported, inductive generalizations.

Discuss this Article (0 messages)