-
@ Econ guy Konrad
2025-05-17 01:49:33This article is a repost, migrating from Substack
Uncertainty is the only constant
It goes without saying that uncertainty is a given in business. It could be the weather, pestilence, or some other physical phenomenon that brings different conditions for your business (good or bad). Or it could be uncertainty in the behaviour of consumers, suppliers, workers, or any other person critical to your success. Most of the time, we just don’t know what will happen next.
One of the most important skills in managing a business, then, is making predictions in the face of uncertainty. We have several tools to do so. The first covers methods from the physical sciences, and involves calculating the frequency of events. We need to unpick a concept at the root of uncertainty, probability.
Thanks for reading Department of Praxeology! Subscribe for free to receive new posts and support my work.
Subscribed
Types of probability
In Human Action, Ludwig von Mises introduces us to two classifications of probability: case and class probability.
Case probability
Case probability is when we know something about some of the factors which determine the outcome, but not others.
This is the kind of probability most often encountered in business. We often know small key parts about several factors that lead to a particular outcome, we might see that oil prices have risen significantly, coupled with a decrease in the exchange rate. These factors help to lead to the outcome of a rising price of fertiliser. But then, there are other factors that lead to this outcome, such as a myriad of decisions by businesspeople all along the supply chain. These, we know absolutely nothing about!
Class probability
Class probability is when we know everything about the particular kind (or class) of event, but we don’t know anything at all about the specific event.
We do encounter class probability in business. But it happens at very specific times (as opposed to just all the time, as with case probability).
Myrtle rust, picture courtesy of NZ Department of Conservation
For example (and sticking to agriculture) we know nothing about an outbreak of Myrtle rust on our farm. But we know, from agricultural science, that the likelihood of an outbreak in our area for a given year is about 13 percent.
So how can we make predictions of these events?
Making predictions
For the latter (class probabilities, like losses caused by Myrtle rust) we can always rely on methods from the physical sciences to calculate frequencies. Then, we know that the physical world has certain constants, so we can say that given a set of conditions (like rainfall) the likelihood of losses from Myrtle rust in future is about 12 percent.
Thinking about the former (case probabilities such as changing consumer preferences, or changes in prices). In these situations we are dealing with the realm of human decisions. So while we can count the number of these events, we cannot use these counts for prediction. The reason is that, in the realm of human decisions, there are no constants. We can’t say that given the world of 2023 prices rose five times, therefore in the world of 2027 prices will rise five times. Even if physical conditions are the same, human decisions will not be.
Luckily, economic science does give us an answer. In dealing with case probabilities we should use our method of understanding (which I wrote about here).
Using this in your business
You can incorporate this knowledge into your business by critically analysing each event for which you need to make a prediction. Is the event dealing with the physical world, for which there are constants? Or is it the realm of human decisions, for which there are no constants? It most likely is a mix of these.
Work backwards and jot down the more granular events which compose the event in question. Then try again to fit each of these into case or class probability.
When each event is broken down into only one category you know which methods of prediction to apply to each event.
For case probabilities, engage a specialist in the area. For class probabilities, engage your own expertise in understanding, and consider hiring an economist to augment your analysis.
