ENGINEERING FOR DEVELOPMENT

8.1 The expansion of economic activity is generated by a multiplicity of individual innovations each designed on the basis of certain data and assumptions and with specific targets in view. Without centralised control of some kind, the design background is selected by the innovator himself (or imposed by his financial backers) usually in terms only of market costs and profitability of the innovation itself, perhaps with some guidance from the State in respect of desirable fields of activity supplemented by tax reliefs or subsidies in specific cases. The inadequacy of such a virtually uncontrolled system in terms of accelerated development is shown by the slow rates of overall economic growth which have been achieved in the recent past. There is no prospect that a large number of innovators acting individually could change this state of affairs even if, with the best will in the world, every one of them accepts the desirability of accelerating economic growth. This applies equally in a "capitalist" society or in a society based on state direction of every economic activity, since the actual innovations in both cases continue to be made in a random manner depending on criteria regarding the use and performance of capital and labour which will effectively be set by individuals. The question of ownership of capital is relatively a side issue compared to the question of efficient use of capital and labour. It is only by injecting into the design of innovations some acceptable polarisation related to the future levels and balances of the economy that their total effect can be made to add vectorially to produce overall effects directed to a common end. As we have seen earlier, these must be twofold: (a) the build-up of consumption through increases in employment and earnings; and (b) high levels of capital formation and very efficient use of capital from the point of view of the economy as a whole.

8.2 Having accepted that some form of central direction is a prerequisite to the achievement of our objective, we can break down the problem into two closely related parts:

Firstly: the generation of a field of new investment proposals which will contain a sufficiently large proportion of individual proposals likely, when operating together, to reinforce the growth needed. This implies that the total volume of proposals coming forward must be larger than the volume needed to meet our requirements so that some can be rejected or postponed. The waste of time and effort in preparing proposals for rejection is only apparent. It is not a real waste, since without an excess supply no flow can be controlled. The proportion of "misfit" proposals can be reduced by setting suitable design criteria which the project designers can accept. These may include requirements which in individual cases will be conflicting, such as: ceiling on selling prices; high return to capital; low investment per job; low content of imported materials. Such conflict in requirements will not bother the designers since they always work under such conflicting conditions; we are merely substituting one set for another.

Secondly: the application of a system of selection to the overall field of projects in order to produce a sufficient number of innovations of appropriate design so that in operation their combined result is likely to maintain the required rate of growth of the economy and the balance of the economy. In this problem it must be emphasised that, both in the design stage and in the subsequent operation of the individual innovations, there will be deviations from one or more of the set targets or criteria. It is unrealistic to attempt to tailor the design and operation of every individual innovation rigidly to the assumed targets or to reject all proposals that will not operate rigidly on target. Variation between individuals is a property inherent in any group of similar objects; it can never be eliminated but its effect on the overall properties of the group can be taken into account statistically.

8.4 These two elements form an "open loop" control system. They are designed to control an input in a way calculated to produce a predetermined output. But we need more than this since the growth we wish to generate is changing both in rate and in balance, with time. It will be necessary for the input controls to vary with time, i.e. by changing the design criteria and the selection criteria in step with the change in the economic situation with time. Criteria suitable to investments becoming productive in YEAR 3 will be different from those in YEAR 1, and both of these must be known during YEAR 0 when projects falling into both classes will be in course of design. These differences can be calculated from the target balances of the economy year by year as set out in Chapters 4 to 7 in the form of "programmed inputs". Even this will be inadequate in long-term practice since the economy as a whole is subject to perturbations outside our control, especially from the climate, and from conditions in other countries. It will therefore be necessary to organise some longer term form of "adaptive control" by the addition of "feedback". This will be based on information about the actual performance achieved in relation to our set targets - overall and sectoral growth rates; balance of foreign payments; balance between production and consumption. Measurement of these is notoriously slow and difficult especially under conditions of rapid change, and it will probably be necessary to set up sampling mechanisms from which relatively rapid estimates can be derived in order to reduce "time lags" to manageable levels. The results would be used to modify the longer term design criteria and also the selection criteria on a shorter term basis.

8.5 Figure 2 shows a simplified block diagram of the circulations in the whole economic system. The two circles in which we are most interested are the "Production-Consumption Circle" in which we aim to increase the volume of flow; and the "Growth Circle" which will induce this increased flow. The main points at which measurement must be made are shown by rectangles and the three essential control points: Growth, Design and Imports.

8.6 The number of economic activities to be extended, expanded or initiated during our three year programme is quite large enough to be treated by statistical methods of control. We can estimate this number if we can find adequate information on the distribution of size of all existing activities, e.g. as measured by the number of persons engaged. The average size is likely to be quite small, certainly under one hundred persons, though individual enterprises will vary from one to possibly five thousand persons. Assuming as an example that in Agriculture the average size is ten and in all other sectors twenty, we start off in YEAR 0 with 39,000 agricultural undertakings and 18,000 others. By the end of YEAR 3 there should be rather less than 39,000 individual agricultural activities (possibly only 30,000) due to amalgamation, and perhaps 20,000 new or extended activities in all the other sectors. The plans for these changes will be implemented by about 10,000 farmers and 20,000 non-farming entrepreneurs and managers. Moreover, for each plan that is implemented, at least one other will have been formulated and discarded. Thus at the design level, the workload can be estimated at the very least to be the preparation of some 60,000 investment proposals in three years.

8.7 One way of ensuring that our design targets are met would be to combine our control system with a central investment planning office equipped to prepare all the required proposals, so that control is integrated with design. This would need a highly-trained technical staff of several thousands who between them cover every discipline and expertise which will be used in every part of the economy (plus clerical and other supporting staffs). Their work would include for each proposal: the collection of background data from the entrepreneur/manager who will ultimately implement the proposal; checking on markets; deciding upon the location and scale of operation; designing the operations, processes, management control etc; selecting and estimating the cost of appropriate equipment; checking the availability of suitable labour, raw materials, supplies and services; calculating the operating costs, selling prices and profits; arranging adequate finance; and "selling" the whole proposal to the entrepreneur/manager. The last point - "selling" the proposal in all its details to the entrepreneur/manager - is the point at which such centralised operations are most likely to break down. Unless the entrepreneur or the person who will be responsible for implementation is in full and wholehearted agreement with every aspect of the proposal, it will not produce the results it was design to produce and the planning effort might as well not be made.

8.8 Such a centralised designing staff would of course be a considerable charge against the government budget. Moreover it would drain off an appreciable number from the already scarce operational technical and business manpower of the country, who could otherwise continue in productive activities.

8.9 The number of investment plans to be designed is of the same order of magnitude as the number of entrepreneur/managers who must implement them. These entrepreneur/managers (whether owners or employees) must be assumed to exist, otherwise the projects will not be implemented and the whole development plan collapses. It must also be assumed that each entrepreneur/manager understands the design and operation of the planned investment, otherwise he will not be able to operate it. Thus, in order to assume that the overall development is attainable, we must assume that sufficient numbers of people exist who can (with some technical assistance perhaps) plan the required productive investment in detail. As noted in paragraph 8.6 there will be some 30,000 of them and their workload in planning will reduce to about one project each, every three years.

8.10 If we can leave the detailed planning to these individuals, three advantages will accrue: firstly, each individual plan will be produced by a person who can be assumed to have the appropriate background of expertise (at least at the practical level); secondly, a lot of the necessary background data is already in his hand, especially as regards markets, location, labour supply, raw materials and services, operating cost data, selling prices and availability of finance. It will of course be necessary to supplement his knowledge through some form of information service especially in regard to growth of markets, improved methods of production planning and control, identification of new products and activities likely to offer openings for profitable investment, and process technology. (Much of the last item - information on process technology - will be provided by the sales personnel of equipment suppliers but this is often unrelated to the level of technology appropriate to the country since it originates in the fully industrialised countries. A local technical information service, biased to the real needs and potential of the country, is always desirable to offset this.) Thirdly: the large number of innovations required will be generated by a similarly large number of individuals, in most cases as a by-product of their normal work and without specific cost in salaries (incidental expenses of planning will commonly be absorbed as overheads in existing activities).

8.11 We will therefore abandon the concept of fully centralised detailed design of all the investment projects needed and assume that they will be generated at the points of implementation. This brings us back to the point made in paragraph 1.4: that uncontrolled innovations generated in this way will be "random" and without the introduction of some "polarisation" that total effect will not be the development we have set out to achieve.

8.12 We come back to the need therefore for a separate central control organisation capable of injecting sufficient "polarisation" into a large number of independently generated investment projects or innovations, so that their result taken as a whole will meet the restraints imposed by the balancing of the whole accelerated economic development which is our target.

8.13 The important distinction here is between designing and controlling. Centralisation of the first is patently undesirable (and probably physically impossible at any except very low rates of economic growth). Decentralisation of the second is impossible since it involves the application of a limited number of criteria (which must change with time) to a large number of proposed investments. The control function involves the application of a single discipline and therefore presents no insuperable obstacle to centralisation.

8.14 The targets upon which a growth control system can be based may be listed as follows:

	YEAR 1	YEAR 2	YEAR 3
TOTAL NEW INVESTMENT BECOMING PRODUCTIVE:
Agriculture Other Sectors, expansion Other Sectors, new activities	$21 m $15 m $54 m	$24 m - $84 m	$29 m - $126 m
Total	$90 m	$108 m	$155 m
AVERAGE NEW INVESTMENT PER WORKER:
Agriculture (all workers): Fixed capital Working capital	$15.5 $38.3	$20 $43	$31 $43
Total	$53.8	$63	$79
Non-Agricultural (new jobs only) Expansion New Activities	$716 $616	- $1010	- $1220
AVERAGE PRODUCTIVITY OF NEW CAPITAL:
Of total new capital in Agriculture Of new fixed capital in Agriculture Of total new capital in all other sectors	1.08 3.70 2.04	1.00 3.15 2.12	1.00 2.55 1.87
AVERAGE EARNINGS PER PERSON ENGAGED:
In Agriculture In all Other Sectors	$440 $660	$496 $743	$567 $835

Using these parameters as "controlled variables" we can influence other measures of performance as follows:

The variable in the above entered as "the drift of workers from Agriculture" needs special consideration and so far we have no controllable variable which might influence this either way. We must therefore take a look in Chapter 9 at the essential elements in the lowest levels of rural activities, including subsistence farming, small holdings and associated cottage industries. Any controls and development incentives which can be applied in this sub-sector will be different from those in the "industrial" sectors, including the industrialised agriculture sub-sector.