Several guest articles in the FAZ recently dealt with the problem of predictability and unpredictability in business administration (BWL).
The contributions by Burkhard Schwenker and Franz Schencking as well as by Klaus Brockhoff represent such contradictory positions that it seems worthwhile to take a somewhat more fundamental look at the decision-theoretical foundations of business administration.
Schwenker and Schencking attest to a tendency towards predictability in business administration, which wants to establish functional relationships in the context of a decision calculation both in the case of safety and risk, which should (inevitably) lead to the optimal decision given the given conditions and objectively determinable probabilities.
Critics objected that the “paradigm of predictability” was linked to knowledge of probabilities and that, in view of the blatant lack of information, in many cases it was necessary to do without this paradigm and instead that several scenarios should be “developed and calculated” (sic!) in parallel.
Brockhoff quite rightly objects that scenario planning is a well-established concept and business administration therefore does not need to be reinvented.
In addition, she has a well-developed arsenal of methods to create predictability even for "events that are perceived as singular and have a profound effect on companies".
Brockhoff refers to the classic “Risk, Uncertainty, and Profit” by the American economist Frank Knight from 1921 and the concept of subjective probabilities discussed therein as a supplement to the objective, i.e. subject-independent, prevailing to date – and according to Schwenker and Schencking still probabilities.
Brockhoff, on the other hand, comes to the opposite conclusion that predictability can always be guaranteed by determining subjective probabilities.
And there is no doubt that business administration, based on the risk-benefit theory according to John von Neumann,
Making decisions under uncertainty has its pitfalls
However, in addition to the concept of subjective probabilities, which can also be found in John Maynard Keynes or Irving Fisher, Frank Knight bequeathed another elementary differentiation to business administration.
It consists of dividing uncertainty, i.e. the state of imperfect information that does not allow predicting the consequences of actions so precisely that only one result is the consequence, into the two subcategories risk and uncertainty.
According to Knight, we speak of risk when information about the probability of occurrence of possible outcomes of alternatives can be given, while this is not possible in the case of uncertainty.
The central question is whether Knight's expansion of the concept of obtaining objective probabilities "either through calculation a priori or from statistics of past experience" to subjective probability assessments ("Estimates") does not completely define the uncertainty situation away.
Because a lack of information about objectively determinable probabilities can apparently always be remedied by subjective assessments.