Abstract
We’ve now identified two different occasions for strategic decision exercises. The first or projective occasion would find strategic decision exercises being used to set the contents of a planning/positioning structure. The result of such exercises would be the specification of an ostensibly best strategic response option for each of the projected event possibilities for which an organization desires to be ready (which may appear as integral models in a contingency planning construct or as generic or template-based referents when it is an adaptive structure that’s being built). The second occasion for a strategic default, such again being triggered by the emergence of an event for which there is no entirely appropriate, pre- defined standing strategic response option (integral, categorical or template-based), such at least some —if not all— of the provisions of an appropriate response will have to be determined on a reactive basis.
Up to this point we’ve taken a rather sanguine perspective on strategic management and decision-making. Particularly, we’ve pretty much assumed that there will be a neat, effectively linear progression through the various requirements posed by the several strategic planning/positioning platforms. But what we now want to do is introduce the prospect of con-foundations that can either frustrate or perhaps even defeat attempts to rationally resolve strategic issues in the real-world. And once we bring these confoundations to light, there immediately follows a recognition of the points at which contemporary management science procedures and instruments can no longer be relied upon to guide us towards rational strategic decision choices. That is, we run up against the prospect of a serious “shortfall” in strategic analysis and decision technology, and hence in strategic management practice as well.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Notes and References
c.f., the classical volume by R. A. Fisher, Statistical Methods and Scientific Inference (Oliver & Boyd, 1956).
The originally-formulated von Neumann-Morgenstem paradigm appeared in their Theory of Games and Economic Behavior (Princeton U. Press, 1947). For a discussion of Marshak’s early refinements, see his “Rational Behavior, Uncertain Prospects and Measurable Utility,” Econometrica (18, 2; 1950).
As used here, uncertainty is taken to reflect the level of ignorance —degree of stochasticity— with respect to actualities, and so might best be measured as a function of the morphology of the preference functions (the configuration of a probability or utility density function). For its part, risk will derive from estimates of the cost/loss (real or opportunity) to which we might be subject should realities fail to align with whatever expectations underlay actual decision choices. This connotation for uncertainty and risk derive from the practicing decision scientist’s focus on a composite criterion, the Expected Value of Decision Error as a basis for evaluating the peril associated with any decision exercise, this being a summary product of the probability that anything other than an anticipated event will occur and the penalty each such possibility would exact.
A very good discussion of this technique is given by Martin Beckman in his Dynamic Programming of Economic Decisions (Springer-Verlag, 1968),
also by Kamian & Schwartz in their The Calculus of Variations and Optimal Control in Economics and Management (North Holland, 1981).
For a treatment of Bayesian-based learning protocols in general, see K. S. Fu, “Learning System Theory,” in (Zadeh & Polak, eds.) System Theory (McGraw-Hill, 1969).
It’s here that perhaps the most pragmatically telling of Maurice Allais’ well-known criticisms of Objectivist computational conventions applies. For he posed the following prospect: “If am able to participate in a long series of games, but could be ruined early on, possibly even in the first round, it is obvious that the justification of the rule of mathematical expectation by the laws of large numbers is invalid” For more on this, see pg. 71 of Allais & Hagan (eds), Expected Utility Hypotheses and the Allais Paradox (D. Reidel Publishing Co., 1979).
Allais & Hagan, Expected Utility Hypotheses and the Allais Paradox (D. Reidel Publishing Co., 1979), pg. 104.
Notable among such formulation are those of Ramsey in his The Foundations of Mathematics and Other Logical Essays (Littlefield and Adams, 1965 reissue of 1931 work), Savage’s The Foundations of Statistics (Wiley, 1954) and Edwards, Lindman and Savage, “Bayesian Statistical Inference for Psychological Research,” Psychological Review (70; 1963).
At least one decision theorist, J. C. Harsanyi, has explicitly considered such a prospect, as per this statement from his paper, “Acceptance of Empirical Statements: A Bayesian Theory Without Cognitive Utilities,” Theory and Decision (18; 1985): “ . . . it is one of the great virtues of Bayesian theory that it clearly recognizes the important roles played both by formal and informal decisions in human decision making. If a Bayesian decision maker wants to find the best policy in a given situation, he has first to choose a specific model for analyzing this situation. Then he has to choose his prior probabilities and his von Neumann-Morgenstern utility function. All these choices are largely matters of personal judgement. But once these choices have been made, finding the expected-utility maximizing policy is a matter of computation and, therefore, represents a formal decision.”
For an important critique of Bayesian epistemology, see C. Glymour, Theory and Evidence (Princeton U. Press, 1980), chapter on “Why I Am Not a Bayesian”.
This fact has led W. B. Weimer to assert that mainstream Bayesian epistemological argument ultimately reduces to “ . . . almost a word-for-word substitution for the confirmation theory of the inductive logicians” (see his Notes on the Methodology of Scientific Research; Hillsdale, N. J., Lawrence Erlbaum Assoc, 1979). It also leads to an obvious technical consequence for Subjectivist-Bayesian driven decision exercises. For it can always happen that a-priori arguments may have virtually no significant impact on the final formulation. As von Winterfeldt and Edwards put it in their excellent discussion of Bayesian-leaming based decision processes (Decision Analysis and Behavioral Research; Cambridge University Press, 1986): “ . . . prior probabilities . . . frequently have but negligible impact on the conclusions obtained from Bayes theorem, although utterly unlimited vagueness and variability would have utterly unlimited effect. If observations are precise, in a certain sense, relative to the (continuous or dense) prior distribution on which they bear, then the form and properties of the prior distribution have negligible influence on the posterior distribution.” And so long as this is can be the case, one might always suggest that the key point in Allais’ famous indictment of the Objectivist formulations of the American school — their neglect of “ . . . the fundamental feature characterizing the psychology of risk, that is the dispersion of psychological values” — may often be of only negligible practical impact. Thus, while this criticism may be an axiomatically interesting foray in the Subjectivists’ axiomatic struggle against the Objectivists, it is not a particularly telling thrust from the working decision scientist’s point of view.
There have been attempts to put such matters in some sort of clinical perspective, a good example of which is the widely-cited essay by Tversky and Kahneman: “Judgement Under Uncertainty: Heuristics and Bias,” in Science (Sept. 27, 1974). But while works such as this provide us with a description of the problem of a-priori prejudice, the tendency is to assume out-of-hand that such are part of man’s natural condition, and so largely beyond removal or repair via any technical or scientific initiatives. However, as we’ll try to suggest in the next chapter, this is hopefully not the case. For there is some promise of being able to devise technical initiatives that can bring matters of perspective, opinion and judgement within the purview of formal apprehension.
There are also at least three ancillary classes of assessment-related confoundations that we might mention: SEMANTIC IMPRECISION, such that one or more of the factors addressed by one or more alternatives is a subjective concept for which there is no generally acceptable exerimentally-accessible or measurable counterpart or surrogate. SYNTACTICAL DEFAULT, this concerned with the tendency for subjective or notional realization constructs (hypotheses) to represent improper hypotheses . . . e.g., they may rest on implicit or transparent assumptions, impose elliptical linkages among their component arguments or be enthymetic rather than properly syllogistic in structure (all examples of logical flaws we’ll be looking at in the next chapter). PRAGMATIC CONSTRAINTS, the most pertinent of which are time and resource strictures. Time constraints are an obvious concern because, outside the leisurely world of basic academic research, the point at which we assessment efforts must be abandoned is more likely to be determined by the pace at which exogenous circumstances are developing than the rate at which uncertainty is being exhausted. Similarly, real-world decision exercises are also generally likely to be faced by a consistent scarcity of analytical resources in terms of skilled personnel, information acquisition budgets, or funds to construct prototypes or establish experimental facilities, etc.
For a fascinating essay on assessing the technical integrity of the Star Wars vision, see Lin’s article: “The Development of Software for a Ballistic Missile Defense System”, in Scientific American (253, 6), 1985.
In fact, SDI is a prime example of one of the luxuries that strategic decision authorities often allow themselves. For as a practical matter, the execution of the Star Wars agenda was undertaken even though the strategic level debate about SDI had not been concluded (nor indeed, had those in power really joined in the debate at all). That is, it was a strategic choice wished upon the world as a matter of simple executive hat. Therefore, at least so far as the Reagan administration and their Pentagon subordinates are concerned, SDI was immediately transformed into an issue of tactical rather than strategic significance. As such, virtually all of the research now being conducted into SDI are intelligible as assessment-related activities directed at evaluating different means to achieve an end that itself was never subjected to any kind of assessment ... exercises aimed at the prospective merits of simple kinetic impact weapons vs. particle-beam lasers, etc.
One of the seldom-mentioned but most pragmatically important factors arguing for these kinds of exchanges has to do with a very disturbing tendency in the general technical community. For increasingly, both academic and professional careers in the technical sector are tied to some narrowly-defined specialty like simulation, game theory, econometrics, cybernetic, Expert Systems or one or another of the programming variants (linear programming, nonlinear programming, dynamic programming, goal programming, etc.). The peril here is obvious enough. For there is always some prospect that a technical specialist may be less interested in actually solving a problem, per se, than in trying define it in a way that will bring it within the purview of the particular method of which he is most fond. And in such redefinitions, it may be that we often let our enthusiasm outstrip our technical discretion.
For an example of the speculative underpinnings of brainstorming schemes, see Ginter & White, “A Social Learning Approach to Strategic Management: Towards a Theoretical Foundation”, Academy of Management Review (7, 2) 1982, or van Cauwenbergh & Cool, “Strategic Management in a New Framework”, Strategic Management Journal (3) 1982. But there is another line of non-technical approach that has much merit, this the attempt to view higher-management from the perspective of the methodological requirements it might be thought to entail. This perspective began largely with C. West Churchman, c.f., his Challenge to Reason (McGraw-Hill, 1968), and sets a tradition that has been carried on most notably by Ian Mitroff and others. See, for example, Mitroff and Mason’s paper, “Business Policy and Metaphysics,” in The Academy of Management Review (Feb., 1982). Or see Mitroff and Turoff’s, “Philosophical and Methodological Foundations of Delphi,” in (Linstone & Turoff, eds.) The Delphi Method: Techniques and Applications (Addison-Wesley, 1975).
As might be expected, the more vocal and tireless exponents of this position are those who have a stake in the Case Method as the basis for executive education. See, for example, Theodore Leavitt’s article, “A Heretical View of Management Science” in Fortune (December 18, 1978). For the case-method approach seeks to prepare budding executives to deal with future problems by exposing them to selected anecdotes describing how their predecessors dealt with situations in the past. The case-method emphasis is thus one tangible expression of the thesis that the essence of managerial training is the distillation of experience (the other popular expression being the principles of management school). There are however two troublesome aspects to this approach. First, it’s not entirely clear how the emphasis on historical anecdotes is to help working executives cope with situations for which there is no appropriate historical analogy ... how, that is, one is to go about generating an original solution to some effectively unprecedented problem? The second point of concern is that if past experience is to serve as the basis for managerial preparation, how does one go about providing for an evolution in managerial sophistication between different generations? Apologists for the case method can of course follow only one line of defense against these criticisms and that, predictably enough, is to insist that management is an art, not a science. Unfortunately, it appears that a great many of those most enthusiastic in making this point are also those least likely to know much about the reach of modem mathematical and statistical instruments, or technical initiatives in any guise.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1989 Kluwer Academic Publishers
About this chapter
Cite this chapter
Sutherland, J.W. (1989). Strategic Analysis Capabilities and Confoundations. In: Towards a Strategic Management and Decision Technology. Theory and Decision Library, vol 8. Springer, Dordrecht. https://doi.org/10.1007/978-94-009-0953-3_6
Download citation
DOI: https://doi.org/10.1007/978-94-009-0953-3_6
Publisher Name: Springer, Dordrecht
Print ISBN: 978-94-010-6919-9
Online ISBN: 978-94-009-0953-3
eBook Packages: Springer Book Archive