Models of Scenario Building and Planning

Facing Uncertainty and Complexity

A. Martelli

• Publisher: Palgrave Macmillan
• Year: 2014
• ISBN:
  9781137293503
• Topic: Business
• Subtopic: Business Strategy
• Index: Business

Models of Scenario Building and Planning offers a unique and innovative exploration of the scenario approach. The book focuses on the analysis of the competitors' behavior; the analysis of risk and uncertainty; and the link between scenarios and strategies.

1

Why Scenario Planning?

1.1 Events, change and trends

It is reported that Harold Macmillan, who was Prime Minister of the United Kingdom in the 1950s and early 1960s, was once interviewed about the political situation. At the end of the conversation the journalist asked him what he feared most for the near future. The answer was “Events, dear boy, events!”

Are events really so fearsome? Sometimes they are. One dictionary defines an event as “a thing that happens or takes place, especially one of importance: the fact of a thing occurring”. The events that Macmillan had in mind when he answered the question were most probably those that occurred in the political sphere and could create problems that were difficult to solve: problems which could divert his energies and those of his government from other tasks, perhaps equally or even more important, but less urgent. Above all, he feared unexpected events against which, by definition, there were no contingency plans. Events of this kind invariably imply change and, in particular, unwelcome change. A change in the state of a particular system between two time points, T₀ and T₁, means only that the two states are different in some respect. Nevertheless, human nature being what it is, the expectation of change always entails some sort of fear, some sensation of impending loss, which can only be dissolved when one knows for certain that such a change is not the harbinger of something negative or dangerous.

But are events of this kind necessarily – that is, by their very nature – unexpected? Not really. In his Discourses on Livy, Niccolò Machiavelli, the 16th century Italian political thinker, claimed that, even if he could not explain how or why, he had noticed that no grave event ever happened in a city that had not been predicted by fortune tellers, by prodigies or by other heavenly signs. In the 21st century, we have largely (but not completely) given up the latter means of prediction, but the need to know something about future events has not diminished; actually, it has increased. So, we strive to devise and to apply methods and techniques which may help us in this direction, even if we know beforehand that results will be, at best, hypotheses to be confirmed or disproved by actual events.

However, one of the reasons (maybe the main one) why events can be better anticipated now than in Machiavelli’s time is that now it is much more possible to place them in a context, and the proper context of an event is the trend of which it is a part. Trends can be defined in many ways, but in essence they are sequences of events concerning the same phenomenon. Trends may refer to a variety of phenomena – in the economic, social or political areas – to fashions, fads and opinions, to production and to consumption. If a trend can be expressed as the variations in a time series (that is, if it comprises a measurable change of a phenomenon in one or another direction), then the anticipation of its evolution can be calculated by means of rigorous statistical techniques. This does not necessarily imply that the anticipation will be correct, as the simple projection of a trend does not include the impact of unforeseeable outside events which might have an influence on it, but it at least provides us with a factual basis to assess it. In any case, if an event is supposed to be the continuation of a known trend, it is certainly easier to anticipate it. In elementary arithmetic, if we have the series 3, 6, 9, 12, 15 and we are asked to indicate which is the next element in the series, we know immediately that it will be 18. On the other hand, even if a time series is greatly extended, in most cases it will only be a portion of the phenomenon under scrutiny and consequently its value will be circumscribed.

Trends are pervasive: they are part of the everyday life of people, of cities, of nations and of the whole of human society. This is the main reason why they arouse so much interest and why anticipating their future evolution prompts so many intense efforts. If we look at the future as a sequence of events which form trends, future trends are the projections of past trends. In this sense, there is an obvious analogy between the analysis of the past provided by historical accounts and the efforts to anticipate the future. The past, too, is a sequence of events of which historians try to make sense by sorting out trends; that is, putting them into some sort of order. And our knowledge of the past is wider than our knowledge of the future, though less than commonly believed. The past is like a map on which we see many points, closely grouped or scattered, as well as many white areas. There is not a great difference between the reconstruction of the past and forecasting and explaining the future, because both require the frame implicit in the reconstruction to be completed – in other words, those white areas must somehow be filled up (Gill, 1986).

The use of trends to anticipate the future evolution of phenomena has sometimes been criticised on the grounds that they are unreliable, are subject to frequent rapid changes and give little evidence of the direction in which things are moving. Basically, it is claimed, they look backward. It is certainly true that trends are useful only in so far as they help in recognising and anticipating change. On the other hand, recognising changes in trends is far easier than, say, comparing events individually to perceive change. No sound theory of change could be construed that does not include trends as one of its basic building blocks.

One important marginal note: changes in historical trends may have unexpected consequences. Taking as an example the trend in the ageing of Western societies (population is actually ageing all over the world, but in the West the phenomenon is much more marked): compared with the data for 2005, by 2025 the median age in Italy will rise from 42 to 47, in Britain from 38 to 41, in the USA from 37 to 39. As a consequence, financial savings will drop – people save less after they retire and younger generations in their first earning years are less frugal than their elders were. The financial wealth of households in the world’s major economies will be roughly $30 trillion less than it would have been if the historical trend had persisted. As these are the areas of the world where the majority of wealth is created and held, the slowdown in global savings rates will reduce the amount of capital available for investment and cut economic growth (Farrell et al., 2006).

Appendix A illustrates a large number of trends, divided into social, economic and political subgroups, held by futurists to be the leading ones of our time. This list is certainly very approximate and incomplete, but it can in any case give an idea of the changes before us in the next 10 to 20 years. Given its limits, the list tells us what problems national and international institutions, governments, religious and political movements, nonprofit organisations and suchlike will have to face, and to what extent tools for analysing the future will be needed.

1.2 Uncertainty ...

Whoever bears the responsibility of deciding on behalf of an organisation knows the paralysing effect of uncertainty, particularly on the evolution of the relevant environment. On the other hand, this experience is not only confined to managers: uncertainty is a leading characteristic of our civilisation. The trends quoted in Appendix A.1 are so numerous and so conflicting as to make certainty about the possible outcome of the general picture an impossibility. And the phenomenon is not new. In economics, according to recent interpretations of his thinking, the most penetrating ideas of John Maynard Keynes (1883–1946), the most important economist of last century, concern the pervasiveness of uncertainty and the part it plays in preventing economies from performing at anywhere near their potential, except at the “moment of excitement” (Skidelski, 2003). Keynes’s main concern was not so much with fiscal fine-tuning as with economic uncertainty, about the cost of capital discouraging investment, for example, thus leading to lower output and to crises.

Another economist, the American Frank Knight (1885–1972), made a groundbreaking distinction: risk means chance when the probabilities of the outcome are known in advance; uncertainty means chance when they are not known (more on this in Chapter 9). If a die is tossed we know in advance that the face with a “5” has one probability in six, (that is, 1/6) of appearing – unless the die is loaded, of course. Unfortunately, most strategic decisions are of a second type: their outcome can only be estimated beforehand on a conjectural basis, where probabilities can at best be estimated, but not calculated. It is indeed arduous to anticipate with some degree of certainty an impending marketing campaign’s chances of success. In a case of this kind, this is usually because the imponderables are too many: in place of a single roll of one die with six faces (each bearing a number), we have an unknown number of dice (each with a varying number of faces and each face with a varying quantity of numbers), about which we know little or even nothing at all. In other words, according to Knight, the idea of risk refers to the cases where probabilities can be measured; uncertainty refers to the cases where they cannot be measured.

This does not mean that we cannot make any educated guess about uncertainty. For example, we know that uncertainty can appear in at least two different forms. The first concerns the nature and characteristics of the factors which determine change. For example, how can the expansion or contraction of a market be interpreted? Or increasingly aggressive competition? Or one technology being replaced by another? Answering questions of this kind is certainly difficult, but not at all impossible if the proper analytical methods, filtered by experience and common sense, are applied. But the task cannot be performed if the first source of uncertainty is compounded by a second, even more dangerous, one. This results from human beings attempting to interpret changes in the environment in the light of their system of values and personal perceptions. Thus, change is not analysed; its explanation is merely entrusted to hopes and visions. The latter can be useful, but they are certainly not sufficient to this purpose. And the quality of the decisions is inevitably badly affected by this second source of uncertainty.

The evidence of the growing importance of uncertainty has led forecasters to look for ways to reduce its impact on the quality and reliability of the forecasts they present. All forecasts, by their very nature, are statements referring to the explicit or implicit probability of the occurrence of a certain variable assuming a certain, usually single, value. Since many single-point forecasts are frequently available for a given variable, if these forecasts show an acceptable degree of agreement, the forecasters confidently expect the outcome they predict to be true. These agreed-upon forecasts, sometimes expressed as averages of the single-point forecasts, are frequently called consensus forecasts and treated as such. The term has entered popular discourse without having been defined in a generally accepted way. But the degree to which an average is representative of the collected individual predictions can vary a lot, depending on the nature of the underlying distribution. The inverse aspect of the consensus is the dispersion of a sample of point forecasts, which can be measured by their standard deviation. Uncertainty in this sense can be identified as the inverse of consensus (Zarnowitz and Lambros, 1987).

Uncertainty cannot be completely eliminated. It can, however, be reduced (that is, transformed into risk), a situation where some ex ante estimate of the chances can be made. In the above sense, therefore, reducing uncertainty can be equated to increasing consensus. Reducing uncertainty is actually crucial to conceiving, framing and applying strategy at any level, be it business, political or military. As such it is one of the fundamental responsibilities of management, and must be accorded a high priority in the process of defining strategy. This high priority makes it possible to make better decisions. Furthermore, any organisation will recognise a decision taken after weighing the chances of the success of an important action; that is, after having reduced the amount of uncertainty. If a decision is not only a risky one, but was also taken without an exhaustive preliminary analysis of the amounts at stake and of the relevant risks, there is no reduction in uncertainty, and sooner or later the organisation knows it. And it either follows grudgingly, or it does not follow at all. Many decisions that in the abstract could be right give bad results in practice.

An effective way to reduce the uncertainty surrounding decisions is to anticipate, by means of rigorous analysis, those changes in the environment which may have an important impact on an organisation. In fact, the anticipation of changes is probably the most effective way, as it stems from the conscious doubt about the capacity of hopes and visions alone to work out what is really happening. In this matter, the role of information and of the skill to manage it is essential.

Traditionally, there have been three different answers to the need to reduce uncertainty by anticipating changes. The first is to throw more resources into the process of forecasting, expanding data gathering and analysis, improving forecasting techniques, using better tools. The second is simply to acknowledge that change cannot be anticipated: the idea is that, owing to the ever-growing complexity of the environment, speculating about the future is simply a waste of time and resources. The third is to accept uncertainty as a fact and to deal with it – not just with one simple anticipation on which to gamble everything, but with a range of possible alternative futures that might arise from the forces of change (Ralston and Wilson, 2006).

1.3 ... And complexity

The changes that impact on the strategy of an organisation stem either from the relevant macroenvironment, outside the industry or the industries where it operates, or from a closer environment, the task environment or operating environment. This includes the entities, events, and factors surrounding an organisation that influence its activities and choices, and determine its opportunities and risks. Changes are usually brought about by an event acting as a detonator. For example, the devaluation of a currency is a signal of a change in the trade flows of the devaluating country, and in this sense it is very easy to interpret. It might be more difficult to evaluate the consequences of a new technological development in a supplying or buying industry. Even more difficult is the case of the appointment at a leading supplier or customer of a senior person endowed with a different mentality from his or her predecessor, and who will most probably follow a different strategy.

The idea of complexity is not attributable to one single theory and perhaps not even to one single, defined and accepted paradigm. Nevertheless, together with its developments, it is nowadays one of the most powerful tools for analysing and understanding how systems, both physical and biological, as well as social and economic, operate. In general, all systems characterised by some principle of self-organisation are complex systems.

The first definitions of complexity were presented in the works of the French mathematician Jules Henri Poincarè (1854–1912), referring to the irregularities present in dynamic systems. But the approaches based on complexity were developed in the second half of the last century, first of all as a reaction against reductionism in natural sciences. Reductionism is the tendency or principle of analysing complex things into their simple constituents. Its basic principle was that natural phenomena are placed within a multilevel hierarchical system with complexity increasing from bottom to top. The more complex levels can be “reduced” to the less complex ones, but not the other way round.

In 1948, Warren Weaver (1894–1978), an American mathematician and scientist who is generally recognised as one of the pioneers of machine translation, made an important distinction between organised complexity and disorganised complexity. The latter is due to a very large number of variables only and is statistically predictable. But organised complexity has to do with patterns that do not submit to the rules of statistics. In organised complexity balance is not due to statistics but to how parts interact. Organised complexity measures this interaction. One important consequence from this is that if in economics we want to use the methods of the natural sciences, then we must give up the idea of equilibrium; that is, that the economy is a sort of self-regulating mechanism. Another consequence is that, as their environment is generally characterised by organised complexity, companies can never satisfactorily deal with complexity. I shall return later in the book to the meaning and significance of these consequences in the use of scenarios.

From the 1950s onward, critics of reductionism expressed growing doubts about its fundamental distinction between truths of reason and truths of fact and the ensuing distinction between synthetic and analytic propositions. Theories contrasting reductionism, from which theories about complexity derived, were considerably strengthened by the ideas presented by Thomas Kuhn (1922–96) in his book about the structure of scientific revolutions, where he first introduced the notion of “paradigm” with reference to what he defined as “normal science” (Kuhn, 1962). A paradigm is a model or a scheme accepted within the scientific community of any given branch of science. When a paradigm meets anomalies, or contradictions, which cannot be solved, normal science usually reaches a crisis point and a more or less prolonged transition begins – the “scientific revolution” – towards the acceptance and adoption of a new paradigm, a process that, as Kuhn asserts, explains the dynamic of scientific progress. With this and other fundamental contributions, the road was opened towards the theories of complexity: it was established that the properties of a system cannot always be...