Flexible Automation in Developing Countries

The impact on scale and scope and the implications for location of production

Ludovico Alcorta

• Publisher: Routledge
• Year: 2003
• Topic: Business
• eBook ISBN:
  9781134653553
• Subtopic: Business General
• Index: Business

This book examines the extent of, and motives for, the diffusion of flexible automation (FA) at global level and then turns to the local and firm level, bringing together in-depth studies of sixty-two firms in Brazil, India, Mexico, Thailand, Turkey and Venezuela. Research focuses on the impact of computer-numerically-controlled machine tools on scale and scope by exploring changes in lot sizes and product variety (product scale and scope), total plant output (plant scale) and total firm output (firm scale). Barriers to setting up FA-based operations are discussed, as are factors which may affect a decision to locate in a developing country. The contributed studies reveal a relatively slow diffusion of FA in developing countries and it is demonstrated that while FA possibly increases scope, it also requires that plant output be increased in order to maintain efficiency. Alcorta concludes that the location in developing countries will probably only be viable for large domestic firms, multinationals seeking to relocate simple but labour intensive assembly processes and firms in countries with significant domestic markets. This work is unique in addressing the scale and scope issues in developing countries and in the wealth of information regarding machine tools which it provides. The data provided in the appendix includes official United Nations data, previously unpublished. This will be of use for all research into trends in the use of machine tools.

Part I
CONCEPTS, METHOD AND SYNTHESIS

1
SCALE AND SCOPE

Concepts and issues

1
Introduction

The purpose of this chapter is to explain the conceptual framework developed in this research and used to address the impact of flexible automation on scale and scope. The concepts of scale and economies of scale and of scope and economies of scope have all too often been used loosely in discussions about the impact of technical change on costs, firm size, industrial organization and industrialization. Much of the confusion has arisen from the fact that the circumstances and relationships these concepts attempt to describe involve a combination of physical, organizational and economic phenomena, which normally are not easy to disentangle. Thus, the first step in this research was to develop a proper understanding of their meaning, of the key factors accounting for their changes, of their different dimensions and of the possible relationships that emerge between them on the basis of established economic theory; and to relate this understanding to the ensuing debate on whether or not new technologies reduce optimal scale and increase economies of scope.

It must be emphasized that standard economic theory, despite its limitations, clearly remains the basis on which empirical research on issues of economies of scale and scope is conducted because it continues to provide a useful and recognizable heuristic instrument to organize the ideas and the data in this field. It also facilitates understanding and comparability across research. It is, therefore, an obligatory starting-point for any research of this kind.

In outlining the development of the conceptual framework the section 2 briefly discusses the origins and development of the theory of economies of scale by examining the contributions of Adam Smith, Marshall and Viner. The main focus is on the intra-firm division of labour, not the inter-firm division of labour, as our concerns extend only to the firm level. Section 3 defines the concepts of scale, economies of scale and optimal scale, and reviews their main sources and dimensions. The sources of economies of scale to be addressed are specialization, indivisibilities and dimensional effects, while the dimensions of economies of scale to be discussed are those of product, plant and firm. Section 4 concentrates on the origins, definition and sources of economies of scope. It begins by addressing Stigler’s contribution, in particular the concepts of plant adaptability and flexibility in the short run, and then moves on to discuss the development of economies of scope theory by Baumol and associates. Section 5 reviews the debate between, on the one side, those who argue that new technologies will result in reductions in optimal scale and economies of scale and increased economies of scope and, on the other, those who contend that the effect will be the opposite. The final section of the chapter develops a graphical representation of both sides of the argument by way of clarifying the debate and as a means of organizing subsequent work.

2
Origins of economies of scale theory: from ‘division of labour’ to cost curves

There are few propositions in economics that have caused so much debate and controversy as Adam Smith’s dictum ‘the division of labour is limited by the extent of the market’ and the pin factory example he used to illustrate his statement. This assertion is seen in both classical political economy and neoclassical economics as one the main foundations of production and economies of scale theory.

In essence, Smith (1776) argued that the principle of the division of labour and the ensuing work specialization was the most important determinant of productive efficiency because of the increase in dexterity which results from making a single, relatively simple, operation the sole activity of the worker; the saving of time arising from not having to move from one type of task to another; and the associated use of machines which ease and abridge labour. In Smith’s celebrated pin factory example, one man draws out the wire, another straightens it, a third one cuts it, a fourth one points it and so on. Up to 18 distinct operations were identified which with no more than 10 labourers could result in thousands of pins produced in one day. A craft workman producing pins by himself would not be able to produce more than twenty pins in the same time. The main outcome of the division of labour was that it reduced direct and indirect labour per unit of output (see also Atallah 1966; Corsi 1991; Morroni 1992; Skinner 1974; Stigler 1951; Tayler 1985).¹

The division of labour was, for Smith, a process resulting not so much from any human wisdom but from the gradual increase in the ‘propensity to exchange’. Because expansion of trade and output is at the origin of the division of labour, the extent of that division is limited by the extent of the market. Where the market is small there is no incentive to specialize as few will be able to purchase the output of those engaged in factory production. Where possibilities of economically transporting products elsewhere exist, the extent of the market is potentially increased by the ‘inhabitants of the lands’ within the reach of those transporting means. For any one country taken in isolation, the extent of its market is limited by its internal transporting facilities and grows ‘in proportion to the richess and populousness of that country’ (Smith 1776:124).²

Classical political economy built on these ideas (Atallah 1966; Corsi 1991; Gold 1981; Morroni 1992; Tayler 1985; Vassilakis 1987). Babbage (1832) added that increasing specialization of labour reduces apprenticeship time and wastage of material. It also helps to separate in a clear way physical and mental labour and, thus, to allocate each job to workers with the appropriate skills and qualifications, resulting in skilled and creative individuals concentrating on activities that require judgement and on the development of new machines and products, and less skilled individuals dedicating to more narrowly defined tasks.

Marshall (1890) brought the link to mechanization more explicitly into the discussion by pointing out that increasing division of labour creates the conditions for the identification and replacement of certain tasks by specialized production equipment. Work that is uniform and monotonous can be gradually taken over by machinery until there is nothing to do by hand except to supply machines with inputs and take away the product when finished. Even the function of overseeing the work of machinery can be replaced by devices which stop movement when anything goes wrong. The main effects of mechanization are to lower the cost of work while making it more accurate because as the division of labour progresses work is continuously subdivided and made highly specialized; and to make possible the manufacture of interchangeable parts.

Marshall also pointed out that ‘internal economies’ arise within any manufacturing firm out of the increasing scale of output which is both the cause and the result of further work and skill specialization, increasing mechanization and improved production organization. Internal economies enable firms to increase their output in more than the proportionate increase of all of their inputs (increasing returns to scale).³

Despite Marshall’s major influence on neo-classical economics and his key contribution to what was emerging as an economies of scale theory, it was Viner who formulated the neo-classical theory of economies of scale as it is known today (Tayler 1985). In his seminal article, published originally in 1931, he developed the first graphical representation of cost theory which included ‘the usual assumptions of atomistic competition and rational economic behaviour on the part of the producers’ (Viner 1953:198). In it he distinguished between short- and long-run cost theory. In the short-run, plant capacity and some production factors such as capital are ‘fixed’, in the sense that they cannot be altered, while others such as raw materials and labour are ‘variable’. Costs associated with each type of factor are, respectively, ‘fixed’ and ‘variable’ costs. Because by definition ‘fixed’ costs cannot be altered, average fixed costs tend to decline as output increases. Since any increase in output is the result of the combination of constantly priced ‘fixed’ and ‘variable’ factors and given diminishing returns for the ‘variable’ factors, average costs will fall initially while the productivity of the ‘variable’ factors rise but will subsequently increase as the productivity of ‘variable’ factors declines. The outcome is a U-shaped total average cost curve. In the long-run all factors become variable and plant capacities were assumed, on equilibrium grounds, to be available to the exact level of minimum average cost output required by any individual producer. Any producer then has the choice of producing at the short-run level of output with the risk of getting it wrong, thus increasing average costs; or of investing in plant capacity believed to match expected demand at any point in the future (hence, the name ‘long-run’ or ‘planning’ cost curve). Because there are limitless options between the short and the long run, and given diminishing returns also in the long run, it is possible to have a U-shaped long-run average cost curve joining all the short- and long-run output and capacity trade-offs—the ‘envelope’ curve.⁴

Viner added that economies of large-scale production were long-run phenomena that only arise out of the adjusting of plant capacity to each successive output. Sources of economies were either technical or pecuniary, emerging out of reductions in the technical coefficients of production or from prices paid for the factors of production. Technical economies arise out of savings in labour, material or equipment due to improved organization or methods of production. Pecuniary economies arise from the possibility of obtaining quantity discounts due to a larger scale of purchases.

3
Scale, economies of scale and optimal scale: definitions, sources and dimensions

Since Viner, textbook scale economies theory has evolved to become a more empirical set of concepts and relationships. It has been influenced also by developments in applied economics and engineering. In textbook production theory scale refers to size of output or capacity of production units, and economies of scale refers to reductions in unit costs due to increases in size of output. Economies of scale are said to exist if total cost rises proportionately less than does output, while diseconomies of scale arise when total cost rises proportionately more than does output; and optimal scale occurs at the point where any increase in output no longer reduces but raises unit costs.

The main sources of scale economies are specialization, indivisibilities and dimensional effects (Morroni 1992; Rosseger 1986; Scherer and Ross 1990). Specialization sources can be further subdivided into static and dynamic sources. Static specialization gains arise out of larger output and had already been identified by Adam Smith in his discussions of division of labour and work specialization. In a nutshell, increasing scale allows the separation of tasks and workers to do their individual jobs rapidly and precisely and with the appropriate skill content, while avoiding the expense of time and effort associated with moving from one task to another. They allow also the use of more efficient purpose-specific machinery and mechanized production processes and an improved production organization. A too-large output, however, may be complex to plan, coordinate and control, leading to management and organizational problems and to losses in efficiency rather than gains. Dynamic specialization gains arise out of the learning potential of long production runs or the cumulative volume of output through time. Where intricate operations and complex process adjustments are involved, efficiency increases as workers learn by doing: the production process is ‘demystified’, for workers and management alike, enabling them to correct for mistakes. Hence, unit costs are reduced as a result of accumulating output.

Any commodity is indivisible if there is a minimum size below which it is unavailable. Morroni (1992) identifies two kinds of indivisibility: economic and technical. Economic indivisibilities arise out of the fact that many commodities can only be traded in a given unit, e.g. (a length of) cloth or (a bushel of) corn. Technical indivisibilities arise out of the physical impossibility of dividing a particular commodity into amounts usable for production and consumption. This is normally the case with capital equipment, the capacities of which vary in discrete quantities and have a fixed minimum. In a vertical sequence of production stages, indivisibilities may arise out of the imbalances that emerge due to the different capacities of capital equipment at each stage as, if all machines are going to be fully utilized, it is the machine with the largest minimum capacity that determines the size of the others. Morroni (1992) also points to labour-related indivisibilities. Wherever there is the need for team work, as in the case of joint lifting of weight, labour cannot be decomposed into its constituents, as it would not be possible to carry out that particular activity individually. Specific skills are available only in certain individuals or groups of them, and often in discrete and minimum quantities. Capital equipment and labour indivisibilities create minimum outlays that have to be incurred by any producer, i.e. ‘fixed costs’. Unit costs fall as any unit of production initially required to produce a smaller output increases its output without a proportional increase in costs.⁵

Sources of dimensional effects relate to the geometrical volume-surface area relationship of certain kinds of capital equipment such as vessels, containers and pipelines. The cost of construction of any container increases in line with its surface area size, whereas its capacity increases with volume. Since the area of a sphere or cylinder varies as the two-thirds power of volume, the cost of building some equipment rises roughly as the two-thirds power of their capacity. Therefore, the higher the capacity the lower the unit costs per unit of capacity. This source of scale economies is commonly known in engineering as ‘the 0.6 rule’.⁶

According to Scherer et al. (1975), Scherer and Ross (1990) and Silberston (1972), scale and economies of scale are better analysed in terms of three dimensions: product, plant or firm. Product scale refers to the volume of any single batch, sometimes referred to as lot size or production run. Product-scale economies emerge from the indivisibility and fixed costs of the operations of preparation of equipment, exchange of jigs and fixtures, machine adjustment and trial runs necessary to begin manufacturing a particular batch or product run (Carlsson 1989a; Kaplinsky 1991; Pratten 1975, 1991a; Scherer et al. 1975; Silberston 1972; Steudel and Desruelle 1992). Batch or lot size is the quantity of identical items or products manufactured in a certain process or sequence of operations between set-ups. Set-ups or setting-up time is the time spent between the production of the last unit of the last batch and the first good item or product of the new batch.

Setting-up times and associated costs are key factors in determining whether and when a new product is manufactured, particularly in discrete product industries (Ayres 1991; Morroni 1991). The classical example is Ford’s replacement of the model-T car by the model-A car, which necessitated closing down the factory for nine months in 1926 (Abernathy 1978).⁷ The car industry has always been under immense pressure not to change car models, which explains why some models remain in the market for years. In the printing industry, the initial typesetting costs can be so high that they sometimes make the publication of specialized or academic books or journals economically unfeasible. Very few books have print runs to match those of the Bible or Porter’s Competitive Advantage. In many operations in the metalworking sector the ‘typical’ set-up time is 20–30 per cent of processing time, while in a US textile factory setting up a roller printing press to print cotton or synthetic fibres takes around 20 per cent of production time (Carlsson 1989a; Scherer et al. 1975).

In general, when production is not on a strict to-order basis, specific decisions on how large a batch size or how long a production run should be are taken with the help of the economic batch- or lot-size model (Mansfield 1988; Steudel and Desruelle 1992; Scherer et al. 1975). In essence, the model seeks to determine the size of the ‘optimal’ batch or lot on the basis of expected demand and set-up and inventory carrying costs, as inventories necessarily build up when large batches are produced:

where

Q=the optimal batch size being sought;

D=the level of expected demand (normally annual);

s=the cost per set-up;

c=the cost of inventory carrying.

What the economic batch- or lot-size model says is that the larger the set-up costs relative to inventory carrying costs, the greater the incentive to continue producing the same item or product if unit costs are going to be kept in check. The lower the set-up costs relative to the inventory carrying costs, the lower the required batch size. As set-up costs approach zero, there is no optimal batch size, and individual batch size can be directly determined as a function of customers’ demand in order to keep low the inventory levels and costs.

Plant scale, in turn, is normally associated with the total output of an entire plant in continuous process or ‘fluid-flow’ industries, such as oil refining, chemicals, steel and cement, in a given period of time—normally one year. It is in these industries where dimensional effects and the ‘0.6 rule’ have the largest impact on economies of scale. But plant scale relates also to the total output or capacity of discrete-good industries, such as printing, mechanical engineering, the electronic, clothing and shoe-making industries. Because discrete-good production involves a wide range of possible combinations of technologies, forms of production organization, and inputs and outputs, the key sources of economies of scale are specialization of labour and machinery, improved production organization, and individual equipment and production process indivisibilities. Both in continuous process and in discrete-product industries, plant economies of scale may arise also out of minimum requirements or indivisibilities of certain functions and needs, such as local management, security, safety, maintenance and the availability of reserve equipment and spare parts, particularly in cases where functions cannot be subcontracted.⁸

Firm scale relates to the output or capacity of the whole firm, which may or may not involve several plants, and firm economies of scale emerge from the indivisible and fixed nature of certain ‘intangible’ investments, such as research and development (R&D), marketing and management. Budgets for the development of new products and processes are normally ‘rule of thumb’ amounts, reached on the basis of previous years’ sales, levels of retained profits, the expenditure of potential competitors, minimum thresh...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• FIGURES
• TABLES
• CONTRIBUTORS
• ACKNOWLEDGEMENTS
• LIST OF ABBREVIATIONS
• INTRODUCTION
• Part I CONCEPTS, METHOD AND SYNTHESIS
• Part II THE COUNTRY STUDIES
• APPENDICES: GLOBAL DATA
• BIBLIOGRAPHY