Century of Endeavour
Technology and the Economy as Interacting Systems
(c) Roy Johnston 1978(comments to firstname.lastname@example.org)
This paper was invoked by Professor Patrick Lynch of UCD as a chapter in a projected book which however never got published as such. I submitted it to Technology Ireland, and forgot about it. There was a change of editors; Paul Hannon took it over in 1977, found it in the pile, and published it in the January 1978 issue. It does represent a reasonable summary of my then current thinking, and on re-reading it I am prepared to stand over most of it. I have interspersed, in italics, a comment or two from a contemporary perspective. RJ 2001.
THIS analysis constitutes a view of the interaction between developing technology and the life-process of the firm. It is a draft contribution to a book on science policy edited by Prof. Patrick Lynch, Science Policy Research Unit, UCD.
It is a view coloured by the experience of the writer, who started productive life as a scientist, evolving towards technology within basic science, making the transition into technology proper and ultimately to a standpoint which can be described as techno-economic by a series of changes of employment.
I begin by characterising the firm, or basic unit of the economic process, as a system or organism which draws sustenance from an environment, deriving from the latter inputs, and excluding outputs, which are the results of transformations carried out by a sequence of logically ordered processes. The role of technology is defined as the dominant factor affecting the qualities of these transformations, and the technological inputs are related to the outputs of another system which exists mainly outside the firm.
I attempt to characterise the technology-producing system, by analogy with the foregoing analysis of the firm, as an organism. An attempt has been made to define the key points of interaction between the two systems, with a view to pointing the way towards the strengthening of the linkages.
The Firm an a SystemWhen making a model of a firm on the computer, it is useful to apply a classification system to the erogenous variables. This is often neglected by people who approach model-building from the point of view of applications of mathematical techniques developed in the abstract. Similarly the endogenous variables require analysis; indeed, it is questionable whether the traditional broad classification of variables into 'erogenous' and 'endogenous' is useful in this context. However, they have been retained as very broad primary classifications.
Thus an 'exogenous' variable may be classified as technological, economic or social; constant, slowly or rapidly changing ;it can be a 'variable' or a 'constraint'; it can be a system variable, an environmental variable, or a policy variable.
These classes are not mutually exclusive. In organising the input for a particular computer programme which models a firm in an environment, the classes of input variable tend to 'make themselves'. The precise structure of slowly changing constraints, environmental, economic, system, technological, policy or whatever variables determines the nature of the interface between the (model) firm and the (model) environment, giving a particular piece of techno-economic analysis a unique flavour.
Endogenous variables fall naturally into two broad sub-classes: variables quantifying the state of the system and variables quantifying its output. At this point, however, it is useful to pause and ask what do we mean by system and environment.
A system may be regarded as a transformation imposed on a set of input variables which generates a set of output variables dependent not only on the input but also on the values of a set of 'state' variables which are embedded in the system itself and which enshrine some or all of its past history.
InadequacyThe inadequacy of the term 'endogenous' is by now quite clear. The initial values of the state variables must be read in as input, along with the erogenous variables. They are therefore, in a sense, exogenous.
Subsequently, however, when the state variables become modified by the various transformations imposed on them, the state variables become endogenous. Some of them, indeed, will emerge as relevant outputs.
Thus, typically, a firm might be specified in its initial state by a set of numbers quantifying stocks, equipment, employed labour-force, overdraft, debtors, creditors, share-capital, etc. These initial state variables would then be modified dynamically by a set of relationships involving, in a unit period of time, variables such as sales (a typical rapid economic variable), production unit-costs (typically these are slow technological variables), prices (often environmental slow variables) etc. This dynamic transformation of the balance-sheet in Period 1 into that of Period 2 is usually described, somewhat crudely, by what is known as the profit and loss account. This may be said to be the embryo of a system-theoretic approach to management accounting.
In a completely deterministic system in a well-behaved or predictable environment, the role of the technologist may be defined simplistically as the 'reducer of the unit-costs'. Within a given system, having a given set of transformations, this is a noble but secondary role. It makes no headlines, if looked at in this narrow sense. Many competent process engineers earn reputable livings in this niche.
Sudden LeapOccasionally, however, a dialectical leap takes place, and someone, either within the system, or outside it in the general system labelled 'technology production' treated below, develops a cost-reduction such that there is pressure to substitute process B for process A. More spectacularly, someone invents a product which transforms the economic environment and makes an old product line irrelevant, or a new line highly relevant and feasible.
Alternatively, some radical change takes place in the economic environment, quite irrespective of technology (e.g. the oil crisis) and it becomes necessary to look afresh at every technological and economic variable and relationship in the system and evaluate them against the possibilities and constraints of the new environment.
The Firm an an OrganizationThe biological analogy implied by the concept of the firm as an organism takes on some real content when one considers the parallels in terms of open-system thermodynamics. Both firm and organism have survival as an over-riding goal. Both take in relatively disordered materials and with the aid of energy build them into ordered systems which aid their survival.
The analogy fails on the output side; in the case of the firm the ordered systems it builds have to supply a utility somewhere outside in the market, otherwise the energy (money) supply breaks down. Thus a firm is an organism which has developed a complete social dimension of which the primary expression is the market.
I have deliberately introduced the terms thermodynamic, entropy and energy because they permit the development of an illuminating approach to the problem of understanding the firm as a system in a stochastic environment. This approach gives some insights into the role of the technologist as distinct from that of the manager. It also suggests a basis for a theoretical approach to management costs, as distinct from ordinary productive unit-costs.
Entropy is a generalised measure of disorder; it is usually expressed as the logarithm of a probability. A reduction of entropy requires an energy input; these two variables are related by a coefficient which turns out to be the temperature. The product of temperature and entropy has the dimensions of energy.
Consider now the economic analogy. Energy, of course, is money; this is no problem. Entropy is disorder: one can define an entropy-like measure of all the various stochastic variables characterising the market, for example. But what is the economic analogue of temperature?
Maxwell's DemonConsider temperature as it is defined in physics in the kinetic theory of gases. It can be related to the average energy of molecules of a gas colliding with each other at random in a box. James Clerk Maxwell invented a concept which will help us here, known to all physicists as 'Maxwell's Demon'. Consider two enclosures A and B separated by a wall in which is a small hole, at which sits a small gentleman equipped with a bat. He is able to see the individual molecules. If he sees a fast one coming from A, he bats it back whence it came. A slow one, however, he lets through into B. With his other hand, he returns slow molecules into enclosure B and lets the fast ones from B through into A. By this means he increases the mean energy of the molecules in enclosure A at the expense of the mean energy of the molecules in enclosure B. The difference in mean energy is measured by a difference in temperature; this temperature-differential is the measure of the ability, or agility, of 'Maxwell's Demon', who stands at the orifice.
This useful figment of Maxwell's imagination gives us immediately the economic analogue of temperature: it is entrepreneurial or management ability.
An alternative way of looking at the ability of Maxwell's Demon is to consider how rapidly he could sort out, say, oxygen from nitrogen; this suggests that his ability could he expressed as a rate of change of entropy.
I gave the temperature-entropy model a preliminary airing in 1972 at the IFORS conference, in my response to Naylor on 'Simulation'. I subsequently aired it again in the 1990s in an article in 'Physics Today'. Both these were somewhat abstracted adumbrations. This 1978 outline here in Technology Ireland is the most comprehensive version of the concept on the public record. I have various unpublished papers in which I have tried to formulate it mathematically, in part by a development of the work of Shannon, but I ran into a mathematical skill barrier. The offer remains open, at the time of writing, to take this up creatively with a theoretician. RJ 2001.
The Model FirmIlluminated by these concepts from the physical world, let us return to our model firm. This may be regarded as taking in material which exists in the environment at a certain entropy level, and placing it in an orderly manner in stock. This is the role of the purchasing management. The production management takes the material from stock, decreases its entropy still further, producing all sorts of improbable and highly ordered items, which are placed in the 'goods outward' store. The sales manager has to look at the disordered market and reduce it to order by finding specific sales at certain places and times, which he supplies from his stocks.
In other words, in the model of the firm we need not only volume-times-unit-cost type terms, but also terms which describe the cost of coping with variability; these terms we have shown should be of the form 'temperature times entropy', which, it will be remembered, has the dimensions of energy and therefore cost.
These temperature-entropy terms, in current management-accounting practice, lurk in the overheads. The analysis of overhead costs is in its infancy; it is beginning to quantify itself as a result of the need to provide economic justification for management information systems based on the computer. The temperature-entropy analogue will perhaps prove to be a useful tool in this analysis.
ClassificationAt the philosophical level it is sufficient to pick out conceptually how the temperature-entropy model of management costs permits the technological inputs to the system to be classified.
Technology can be used in two quite distinct ways:
1) to reduce the entropy level at the input.
In other words, it is possible to impose quality or delivery criteria on the raw materials so as to ease the task of the management of the physical inputs. Alternatively, it is possible to accept a given entropy level at the input and increase the ability of the input management to deal with it, reducing it to the necessary state of order.
Similarly, in the production system, it is possible to spend money on reducing the inherent variability of the material, thereby hoping to reduce overall production costs, or alternatively to reduce the cost of coping with the variability by developing a more flexible production system. These may be labelled the 'quality control' and the 'process control' approaches to reduction of production costs. These, of course, are not mutually exclusive.
Entropy reductionFinally, in the sales system, it is possible to reduce the entropy of the market by creating controlled demand through advertising, or to accept a high-entropy market and develop a rapidly-responding flexible system, carrying large stocks if the production process is relatively inflexible.
It is necessary to remark that the temperature-entropy model of management costs suggests that there is an optimal level of entropy within the firm which is not zero. In other words, it is possible to overdo expenditure on entropy-reduction, putting men of ability in impossible situations, ending up with a high entropy, high temperature situation at an astronomical cost.
Looking at the same thing in an alternative way, given that all decisions are made in a fog, and that sometimes by analysis the fog can be cleared slightly, how much analysis should be done and how quickly? With the workforce available the analysis can be done in (say) a month, leading to a clear indication what the decision should have been a month ago. Quadruple the workforce and get an answer in a week: is this necessarily better? Install a computer and have the answer in a day: is this better still? Without being specific it is impossible to say. We can, however, he certain of one thing: to have complete knowledge at decision-time is infinitely costly.
It is perhaps legitimate to conjecture that a study of the temperature-entropy terms in a system of large organisations engaged in planning against each other might give some insights into the reasons for the general cost-escalation which underlies the abstract label 'inflation'.
The Technology Production SystemThe system which is responsible for the supply of new technology to industry may be looked at in two ways:
(a) as a source of trained manpower.
The system itself consists of various sub-systems:
(1) the Universities
This system, in the UK and Ireland, is rather loosely structured. Indeed, it is questionable if it can be considered as a system, 'within the meaning of the Act'.
Thus, the supply of manpower is ensured by sources (1) and (2), while the supply of solutions to problems is ensured mainly by sources (3) and (4).
In a competitive economic situation, however, the only real source of solutions to problems for a firm is its in-house ability. In general, firms are reluctant to go outside, except for generalised background information; this they may tend to get from Research Associations or Applied Research Institutes relevant to their own industry, rather. than from the Universities of Colleges of Technology.
The transfer of solutions to problems across the boundary between a firm and an Applied Research Institute, given that a firm has been courageous enough to go outside its own R & D unit, is fraught with problems, in that in general it is not associated with mobility of manpower. Industrial technology is full of case-histories of prototype developments done outside the firm which have not survived the grafting in.
RecruitsManpower from the educational system, in the form of raw graduates, is the primary input into both Applied Research Institutes and industrial R & D units. This material is 'raw' in that it is usually unfamiliar with the problems, or even with the basics of the technology. It has to undergo lengthy in-house re-training. Some industries have a tradition of in-house training so strong that they are virtually independent of the third-level educational system.
Clearly we are dealing with a high entropy system, which is loosely coupled to industry, and which has a rather long response-time. We are in a situation where we have a choice between trying to reduce the entropy-level of the raw material, or reduce the cost of reducing the entropy in-house.
Any attempt to do the latter implies setting up an in-house educational system, at an intensive level, which is intrinsically a high-cost operation. To attempt to increase the efficiency of an in-house educational system would be a laudable goal, which if achieved would presumably he applicable in the educational system in general. However why should firms have to spend money doing this? All the arguments seem to point to the need for the firm, or the industry, to have influence over the numbers and quality of its recruits, and to guide consciously the investment decisions of the State with regard to the third-level educational system.
Consider now the third-level educational system as it has evolved. It consists of a number of specialist disciplines, together with various professional schools such as medicine and engineering which make use of the services of the specialists. The latter have developed honours degree courses and basic research programmes.
There is evidence that the numbers emerging from these honours schools are not determined by market forces at all, but by fashions, under the domination of
(a) areas of scientific technology which can attract State money because of long-term military considerations.
Hard core of problemThese honours schools produce graduates having a spectrum of abilities; the top ones tend to be re- absorbed into the academic system, perpetuating its traditions, while the lower two-thirds tends to 'drop out' into industry, often with a sense of failure, and usually with a training which is ill-adapted to the needs. This is the hard core of the high-entropy raw material problem, as seen from industry.
There is a strong resistance on the part of academic scientists to responding to the needs of industry. They feel that the State and industry owes them a living, and that they should be free to choose their research priorities. This is sometimes referred to as the principle of academic freedom. For as long as this principle dominates the educational system, it is unrealistic to expect a tightly-coupled system to develop which relates manpower to needs.
At this point it is necessary to probe deeper and examine the question of social responsibility. Clearly ones attitude to academic freedom depends on the situation. There are two extreme situations, and an intermediate one in which Ireland finds itself. The extremes are:
(a) industry is under conscious social control, by means of novel and as yet imperfect state structures which are being 'de-bugged' under somewhat difficult conditions in the various Socialist countries (eg USSR);
(b) industry is under the control of private corporations responsible to shareholders rather than to society as a whole.
At this time I was clearly aware of the problems of top-down centralist bureaucratic planning as practiced in the USSR, but I had hopes that with the easing of the 'cold war' and the opening up of interactions, the situation might be remediable, and eased in the direction of co-operative ownership with elected boards of directors, responsible to those concerned across the total social spectrum, not just the shareholders. Something like this might have evolved out of the Horace Plunkett process, as espoused by JJ in the 1920s. Unfortunately the bureaucratic factors in the USSR proved intransigent and democratic reform when it came was aborted by various mafia-like process emerging from the earlier pathology. I treat this process elsewhere.
State InvolvementIn the first case, provided the system visibly works in a socially responsible manner, an academic need not worry unduly about being tightly coupled to it. Indeed, the amount of money being spent on very basic research in the USSR, despite the tight coupling to industry, suggests a high degree of long-term social responsibility at State level.
In the second case, academics have every reason to fear the harnessing of the third-level system to the satisfaction of short-term business demand for specific types of manpower, and their clinging to the principle of academic freedom is in fact an instinctive democratic defence-mechanism.
I had seriously underestimated the extent of the corruption of the system. They had been doing good science, but in isolation from industry, and using technology dominated by the military-industrial complex. Post-Soviet Russian science is now almost third-world status, and is decimated by brain-drain. But the basic argument holds: the key is democratic accountability and social responsibility on the part of the State, and this is incompatible with subservience to trans-national corporations.
Separate categoryIreland, however, fits neither of these categories. Native industry is not dominated by a military-political complex; it is small scale and has a survival problem in the undergrowth of a jungle dominated by the multi-national corporations. The development of an effective tightly-coupled system for relating the colleges to industry by a process of technology and manpower transfer would be a positive act, helping Irish industry to survive by reducing the entropy-level of its supply of human material.
A small national economy can only survive in the multi-national jungle by adoption of socialist-like policies; this has long been realised by successive Irish Governments which have on the whole a positive record of state enterprise, despite a basically conservative political position.
A Model Technology Transfer SystemThe following institutional model is one which would work at its best in a situation where all firms were under social ownership and competition did not exist in the capitalist sense. In other words, an innovation made by one firm could rapidly be extended to all, without obstacles.
This is a situation towards which Irish industry, by adopting a co-operative defence strategy relative to the multi-nationals, could evolve, given some State encouragement.
The proposed tightly-coupled system would take the universities, colleges of technology and applied research institutes and build them into a single system which produced both graduates and solutions to problems. The sponsors of the problems would be the in-house R&D units of the firms; the consultants to whom the problems were put out would be closely associated with the training of graduates in relevant technology, by reason of their being also on the academic staff.
The trainee graduates would train, by a quasi-apprenticeship procedure, on problems arising in their place of eventual, or indeed present, employment arid the system would lend itself to release for re-training of mature industrial R & D people.
Financial levyThe basic specialist disciplines would maintain their service role, drawing their funds for independent research by a levy on the turnover of the front-line technological sectors which they serve. Thus basic scientific research would be fostered by a means which allied its interests to that of the technologists, instead of competing with it as at present.
There would be mobility of staff between the technology and technologist producing system and the in-house industrial R&D system. Thus teaching staff would return for a refresher spell in industry, while industrial R&D staff would have the option of teaching, without loss of status, pension rights or any of the artificial obstacles which bedevil the present system.
The career profile of the average scientist would consist of a series of options whereby he could step from science to technology to management and back, from research to development to implementation and back. It happens that this profile fits my own personal career, which has been rewarding, intellectually but hazardous financially. If the financial disincentives were removed (the principal one being loss of pension rights on change of job outside circumscribed categories) this practice of mobility would help along the development of a tightly-coupled educational-productive system under responsible social control.
Some navigational notes:In general, if you click on the 'Back' button it will bring to to the point of departure in the module from which you came; this may be different from those listed at the bottom. Even if the module from which you came is listed at the bottom, it is still preferable to use the 'back' button, because the explicit bottom-of-page hot-spot will return you to the beginning of the module, and you will have to scroll down to get back to where you were.
Copyright Dr Roy Johnston 1999