The role of pecuniary external economies and economies of scale in the theory of increasing returns

Abstract

This paper investigates some issues relating to the phenomenon of increasing returns: (1) What is the role of economies of scale in the theory of increasing returns? (2) Do pecuniary external economies lead to market failure and justify intervention in the market mechanism? (3) Are increasing returns sector-specific or generalised, and if they are sector specific, is it possible to identify and promote these sectors from a policy point of view? We argue that economies of scale are incidental to the broader phenomenon of increasing returns and therefore cannot adequately explain their existence. On the second question, we argue that the presence of pecuniary external economies is characteristic of a well-functioning market system rather than an indication of its failure. Finally, increasing returns are generalised, so that policies intended to identify and promote specific sectors will tend to distort intersectoral relationships. Sector-specific polices should not be based on the logic of increasing returns, but should aim to correct sector-specific handicaps.

Acknowledgment

The authors are grateful to two anonymous referees for helpful comments and suggestions.

Notes

¹Note that in the Youngian conception the emphasis is both on existing as well as new knowledge (or technology) as opposed to the more recent endogenous growth literature, which emphasises new knowledge only.

²In the Youngian conception, there could be two ways of looking at the concept of pecuniary external economies. Firstly, since they lie outside a firm and are all pervading, they can be viewed as the main source of increasing returns to the whole economy. Even the economies achieved by a firm through large-scale production, specialisation and innovation are greatly influenced by them. Alternatively, since they operate through the price mechanism in the form of reduced costs and prices, they can be viewed as a vehicle through which increasing returns are transmitted.

³The distinction is explained in more detail in Section 2.

⁴‘Complementarity of different industries provides the most important set of arguments in favour of a large-scale planned industrialisation’ (Rosenstein-Rodan, 1943, p. 205). In this context, Rosenstein-Rodan regarded the bulk of wage goods as complementary.

⁵The ‘big push’ could be undertaken either directly by the state, aided by a huge inflow of foreign aid; or alternatively, if the private sector is involved, then the price mechanism would need to be supplemented by additional signalling devices to arrive at optimal results, particularly with regard to the size and composition of investment. In both cases, the market is viewed as an unreliable device.

⁶Thus the idea of a government sponsored ‘big push’ seems to have come full circle. Starting with Rosenstein-Rodan (1943, 1961), Nurkse (1953) and Scitovsky (1954), the idea was lost in the neoclassical reaction against what were widely judged to be misguided government interventions of the 1950s and 1960s. In his 1993 paper, Krugman hoped to launch a ‘counter-counterrevolution’ to resurrect the core of earlier development theory by putting it in a more rigorous mathematical form.

⁷‘No regulation of commerce can increase the quantity of industry in any society beyond what its capital can maintain. It can only divert a part of it into a direction into which it might not otherwise have gone; and it is by no means certain that this artificial direction is likely to be more advantageous to the society than that into which it would have gone of its own accord’ (Smith, 1776, I, p. 475).

⁸The function F should include not only outputs and inputs but also their prices, as profits are meaningless without price information. Moreover, in both the functions, external economies require not just that the elements to the right of the semicolon be positive, but also that the partial derivatives of f or F be positive with respect to these elements. Scitovsky seems to overlook these points.

⁹Note that this definition differs from what Young had in mind. Here the emphasis is microeconomic whereas Young's emphasis was macroeconomic. See Note 2.

¹⁰This, however, does not imply that there is no empirical case for increasing returns to scale. At least for certain industries, such as sugar and steel, scale economies may be quite important. For example, Tribe & Alpine (1986, 1987) found that they were significant in the sugar industry of some developing economies. But the important point to be emphasised is that the scale at which a firm is able to operate is itself adjusted to the overall market and cannot be regarded as independent of it. Certain roundabout methods of production become economical only when their advantages can be spread over the whole industry (or economy). Thus, as Young (1928, p. 539) notes, the scale on which firms and industries are able to operate ‘is only incidentally or under special conditions a matter of the size of the individual firm’ and ‘merely reflects the size of the market for the final products of the industry or industries to whose operations their own are ancillary.’

¹¹In the Youngian vision, inputs are seen as an effect rather than a cause of growth. Considerable evidence exists to suggest that growth is not caused by inputs of labour or capital. For a summary of this evidence in the case of the US, see Currie (1997). See also Blomström et al. (1996) and Chandra & Sandilands (2003) for evidence that capital accumulation is not a cause but an effect of growth.

¹²It is true that some advantages may result from the market process itself, but as new firms enter the field they are likely to disappear. However, in cases where monopoly power confers persistent unfair advantages on some firms, the state can intervene to regulate monopolies or ensure more competition.

¹³‘Large production, not large scale production, permits increasing returns’ (Young, 1990, p. 54).

¹⁴Smith pointed out that nations differed from each other not so much in the state of their agriculture but in the state of their manufactures: ‘The most opulent nations, indeed, generally excel all their neighbours in agriculture as well as in manufactures; but they are commonly more distinguished by their superiority in the latter than in the former … . In agriculture, the labour of the rich country is not always much more productive than that of the poor; or, at least, it is never so much more productive, as it commonly is in manufactures’ (Smith, 1776, I, p. 10).

¹⁵However, Smith was not a dogmatic laissez faire economist. He advocated self-interest as long as it promoted public interest. When self-interest conflicted with public interest or when public interest was inadequately promoted, Smith favoured intervention. For example, he advocated public education to the poor ranks of people to counter the ill effects of the division of labour. He advocated public health measures to eradicate disease. He favoured military training to provide for national defence. To enlarge the size of the domestic market he favoured public investment in transport and communications (along with free trade). He also wanted the legal rate of interest, in those countries where a maximum rate is fixed, to be a little above the lowest market price, to channel credit to prudent borrowers (who can make profitable use of funds) and to discourage prodigals and projectors. See Viner (1928) for a long list of functions which Smith believed the state could legitimately perform, depending on the merits of the case.

¹⁶In his review of Pigou's Wealth and Welfare, Allyn Young (1913) suggested that in increasing cost industries the rise in costs is not due to use of more resources but to bidding up of prices of specialised factors as the output expands. So there is just a transfer of purchasing power from industries using specialised inputs to the owner of factors. Pigou (1920, p. 798) at first tried to refute Young's contention in the revision of his book, retitled the Economics of Welfare, but later corrected his error in the 1924 edition after additional criticisms from Frank Knight and Dennis Robertson. See also Blitch (1995, pp. 38–39).

¹⁷The conception of competition in Smith is very different from the neoclassical idea. While Smith viewed competition as a process, the neoclassical conception implies an end result such as Pareto-optimality. Moreover, in Smith the division of labour arises in exchange, so institutional arrangements facilitating market exchange are desirable. The state should therefore ensure two things: first, each individual's liberty to pursue his own interests, subject only to the laws of justice, and second, a secure climate for the accumulation of private property. For details of Smith's institutional approach see Chandra (2004). In the neoclassical version, the end-state conception of competition appears devoid of an institutional setting (see for example Hodgson, 1999).