vii The gazer into the future has never yet found a really comfortable intellectual position, and perhaps never should unless, that is, he is, a preacher.

vii Most forecasts are wrong.

[JLJ - But is it absolutely necessary for a forecast to be correct? Ebenezer Scrooge viewed a forecast of his future life and changed his present behavior. The way it turned out, the initial forecast was wrong. Perhaps what we are really after is a vision of the consequences of our present behavior, should it continue without change. The forecast then allows us to change our behavior, if we choose to do so.]

viii To be rational in this context means making explicit the forecasts which underlie decisions so that they can be examined and discussed, and in the light of that scrutiny coming to a judgment about the probability or... the improbability of any of the forecasts being borne out. We are dealing with uncertainties, but that does not mean we should rely only on hunch any more than we should rely only on systematic forecast. To decide what is the proper combination of hunch and forecast for himself... and in any particular setting... is a decision any good decision-maker has to make.

viii four propositions... first, that all decisions are decisions about what should happen in the future; second, that all decisions are based at any rate implicitly on forecasts of what will happen in the absence of any decision, and in the event of a variety of possible decisions being taken; third, that on the whole the more thought that can be given to forecasts in the time available before the decision has to be taken, the better; and fourth, and above all, that however much thought is given to forecasts they should always be treated with scepticism.

Forecasting and the Social Sciences, Michael Young, p.1-36

p.2 it has become accepted that decision-makers... have a duty to cope with changes and control them. This they can hardly do unless they make an attempt to foresee them.

p.2 if the future is to be charted, guesses have to be made about the decisions that policy-makers will take.

p.3 [a] competent decision-maker... His objective is flexibility. The more that can be done to preserve flexibility, the more is his power of initiative retained and the greater his capacity to adapt to changing circumstances. But forecasts cannot in practice be avoided. Every time an investment is made an implicit or explicit forecast is made about its yield over a term of years to come, and the same goes for practically all other decisions whose consequences will obviously spill out into the future.

p.3 the competent administrator is bound to make forecasts... and yet at the same time to arrange that the forecasts shall be continuously or periodically recast in the light of changing circumstances. A budget... is the simplest example of a forecast which is a guide to action but one which is adjusted as events alter the assumptions on which it rests.

p.9 Mistakes will not become all that less prevalent; their inevitability is one reason for expecting demand to increase. Another is that any serious forecasting generates further demand for the same thing to be done by others. The process is cumulative.

[JLJ - Forecasting is merely one requirement of a business plan or a personal plan for the future. You cannot "outforecast" anyone and receive any practical benefits from the accomplishment without the necessary follow-on steps of being strategic in taking a present stance and in subsequently modifying that stance. A forecast does you no good if you do not have the resources to seize the initiative and behave accordingly.]

p.13 the past... is the sole guide to the future as much for the social scientist as for anyone else, and at root the 'reasoning' employed by each is not so different. Extrapolation in some form is used to begin with by almost everybody, if with varying degrees of finesse. In its crudest guise the supposition is that the future will differ from the present in the same way as the present differs from the past. This simple idea exerts a very strong sway over almost all minds.

p.15 The more that past and present interrelationships are understood the more discriminate their extension. This is partly why short-term economic forecasts are relatively so refined.

p.17 Forecasting can be of value to administration - that is why we expect demand to increase. Can it also be to the social sciences?

p.17 a social scientist... If he ventures to predict, he knows that his ability to do so convincingly depends upon the data he has about the past and present. To know where we are going, we first have to know where we are and where we have come from. To do his job better he needs more and better data.

p.20 The purpose of science, either natural or social, is explanation, not prediction. But prediction is one of the best means of testing explanation, although not the only one.

p.20 It is only when some regularities can be discerned in the process of change as they have happened in the past that one may be able to guess a little more confidently.

p.21 One [approach to forecasting] is to scan the fluctuations in any particular variable or set of variables in search of a pattern. The simplest is a cycle, that is a fluctuation upwards and downwards over time which is both regular and recurrent. The relevance for forecasting is obvious. If, oscillating from one limit to another, events did follow an orderly sequence, then once the present position in the cycle were known the succeeding position could be forecast with some rigor... The trouble is that the fluctuations seldom do dispose themselves in such an orderly way.

   Yet sometimes they do.

p.25 forecasters have available to them a range of 'leading indicators', that is a series whose turning points frequently come before turning points in other series. The leading indicators sometimes give false signals or oscillate too much to act as guides. But they are still of some value to forecasters, when used along with the other information they have available.

p.29-30 as I said at the beginning of this paper, sensible men do not forecast unless they have to, and when they have to, they retain flexibility and allow as far as they can for things turning out differently from what they expect. But the more choice there is, and the more 'decisions' that have to be made, the more often people have to commit themselves to a specific forecast about the consequences which will attend their decisions.

Concepts, Methods and Anticipations, Fred E. Emery, p.41-70

p.44 In populations of living systems each is part of the environment of the others and hence they constitute together a social field. If we take this social field as a superordinate system, then the first and most general question is: 'What are the system-environment relations that typically determine survival in this superordinate system?' It should be noted that this question is not directed to each part in turn, but is directed to all parts at once, both as systems in themselves and as environment for the other parts. Answers to this question should tell us, if only in a very general way, the challenges and possibilities implicit in the future development of the system.

 To my knowledge there has been less explicit analysis of another vital question: 'What are the tendencies in the system toward generating the conditions that make adaptive survival behaviour possible?' It seems that if there is to be survival in a particular environment, value should be placed on certain kinds of preparatory behaviour at the expense of others, and changes in the conditions of survival should induce changes in these values or goals.

p.45 Another way of answering the question about survival behaviour is to identify the starting conditions (coenetic variables) that have arisen from past adaptive responses and act as a constraining and guiding influence on subsequent behaviour.

p.49 The earlier the detection, the greater time available for response.

p.54 Predictions are more likely to be valid if derived from analysis of the broader systems.

p.59 When a situation becomes too complex for organized meaningful learning, an organism regresses to vicarious trial and error behaviour - it responds first to this and then to that in a way which is unrelated to the structure of the environment but may be highly correlated with its own prejudices.

p.63 What makes choice unavoidable is what we might clumsily call a design principle. In designing an adaptive self-regulating system, one has to have built-in redundancy or else settle for a system with a fixed repertoire of responses that is adaptive only to a finite, strictly identified set of environmental conditions.

p.65-66 Systems management... Its characteristics clearly relate it to the general problem of environmental transformation... (6) an increased capability of predicting the combined effect of several lines of simultaneous action on one another; this can modify policy so as to reduce unwanted consequences or it can generate other lines of action to correct or compensate for such predicted consequences (Way).

p.68 Hirschman and Lindblom have spelt out in some detail the characteristics of policy-making under these conditions of environmental complexity, uncertainty and value conflict. Our own detailed studies of the decision processes in large-scale systems lead us to agree with them that these processes are most effective when they allow for the co-ordination that arises from the mutual adjustment of the values and interests of the participants even though these social processes may not be consciously directed at an explicit goal, and the decision processes are characterized by disjointed incrementalism.

Future Research Needs of Transport Planners, Christopher Foster, p.71-82

p.72 if investment decisions are to be taken rationally, they must look far into the future. Technical innovations may make some assets obsolete and the normal processes of investment appraisal must always weight future years less than current... interlocal models which are often complex in nature as well as requiring a great deal of information are needed to explain the interactions of any one locality with all others... factors which may normally be ignored in economic decision taking, because in the short run the effects they produce are small, cannot for such long-term decision be safely ignored.

p.73 one cannot predict how complex a model will be needed to explain the as yet unexplained but the tendency is swinging away from simplicity.

p.82 The importance of formal models, which are currently being developed... in increasing numbers and with increasing complexity, is that they allow the effects of changes in various factors to be examined, without any commitment as to whether or not such changes will occur - in other words to test for the sensitivity of final outcomes... Judgment, then, will continue to determine what priorities should be, and judgment tends to be influenced by short-term pressures.

Physical Planning and Social Change, David Grove, p.83-94

p.83 Because of the difficulties of predicting anything it has become a commonplace to say that physical plans must be 'flexible'... Obviously, plans cannot be infinitely flexible. Indeed, once proposals have been embodied... the limits of flexibility may become quite small. But some flexibility is still possible and its range might be increased by suitably planning and building techniques.

Notes on Social Forecasting, Bertrand De Jouvenel, p.118-134

p.119 The most sophisticated forecasting at present is done for short-term changes in the economy: it is based upon an elaborate representation of the relationships obtaining between numerous features of the economy in the form of a system of equations tying together a great number of variables.

p.120 The purpose of the model is to guide policy-makers... the model is itself an instrument for policy-making... the short-term forecasting suggests the policies which will prevent developments from getting out of line, medium-term forecasting suggests the policies which will procure the best pace prudently available... the long-term 'prospective sketch' ...takes into account intervening policies... To put it shortly, momentum policies are taken for granted while the sketch is used to examine the problems arising from change... any so-called 'prediction' is always a starting-point for examination of what should be done... A prediction is properly an expectation which is offered and accepted as reliable.

p.123 This suggests a mixed strategy, which on the one hand seeks to deflect, on the other hand allows for the probability that the deflection will not be successful... But who is this 'you' to whom I am presently recommending a mixed strategy? Obviously it is the public decision-maker, to whom the forward-looking expert addresses his warnings, recommendations and appraisals of potential effect.

p.123 If we seek to exercise some control over the future it is because we hope thereby to make it better than it would be otherwise. We must then use, whether explicitly or implicitly, some criteria of what is 'better'.

p.124 What we want, when elaborating a long-term policy, is to put our typical citizen in a better situation than that resulting from any alternative we can think of... 'The heart of the matter lies in deciding what benefits should be included and how they should be valued' (Dorfman).