2.1 Knowing and Acting

Heterogeneity of individuals is recognised in mainstream analysis in terms of differentiated preferences for goods and differentiated skills for workers. This heterogeneity is pointed to as justifying a downward slope to consumer demand for individual products and an upward slope of supply for labour with specialised skills, both contribute to rationalising the stability of equilibrium. However, no heterogeneity is allowed in how individuals make decisions, with every decision being an optimising decision usually based on perfect information or at least rational expectations. Also, whatever heterogeneity of individuals exists is assumed constant for short-run analysis and determined by external influences in long-run analysis.

Imagination is central to the creation of novelty. As Shackle (Reference Shackle1959, p. 753) notes, ‘in a nondeterministic universe where creation of something essentially new can happen from moment to moment, then the individual imagination seems to be the locus, so far as human beings are concerned, of this continual projection of essential novelty into the world process.’ When faced with choice, individuals may respond creatively and act in novel ways.

Human beings are by nature inquisitive and to the extent that they think differently can conjecture different answers to the same question. These different answers transform the state of knowing, such that every solution to a problem has the capacity to define further problems and the growth of human knowing becomes autocatalytic and open. Because individuals know differently, it is not surprising that they behave differently and their behaviour changes over time as their knowledge changes. Loasby (Reference Loasby1999, p. 43) highlights the contribution to economic evolution from heterogeneity and continual change in individual knowledge and behaviour,

Deciding and acting in the presence of uncertainty are problematic in modern economies (Levine Reference Levine1997). The certainties of traditional society have been removed and replaced with an environment subject to the vagaries of restless knowledge. No matter how carefully we develop knowledge, we may be surprised by an unexpected outcome as most action takes place in an environment of at least partial ignorance. Every action generates new information and potentially leads to a change in knowing. Loasby (Reference Loasby1999, p. 149) references Shackle in noting, ‘the incompleteness and dispersion of knowledge are a constant source of opportunities for creating new knowledge; as some ambiguities are resolved, more are revealed, and people are inspired to imagine new ways of closing their cognitive systems’.

The occasional discovery and creative action occur against a background of routinised behaviour. Uncertainty, limited cognition, and imperfect information encourage behaviour according to habits or routines. Hodgson (Reference Hodgson1997) argues habits and rules are ubiquitous in governing individual decisions. Both habits and rules ‘have the form in circumstances X, do Y’, where ‘Rules do not essentially have a self-actuating or autonomic quality but clearly, by repeated application, a rule can become a habit’ (Hodgson Reference Hodgson1997, p. 664).

Habits and rules are in part based on personal experience, so reflect learning from both expected and unexpected experiences (Witt Reference Witt2001). They are also partly based on the social and institutional environment in which the individual operates. The role of institutions and society in formation of habits is at the centre of evolutionary theorising about individual and household behaviour by Veblen (Reference Veblen1899). An individual acts in accordance with their perceived and desired position in society, leading to patterns of consumption that include conspicuous consumption, bandwagon effects, and snob effects.

Individuals are always constrained by rules of the game governing their freedom to act. In a market economy many informal restraints that shape interaction arise in the context of local interactions, local knowledge, and the specific experience of time and place, leading to geographical and historical specificity. Governments set regulations governing the way market interaction occurs (dispute resolution, quality assurance, health and safety matters for example). These regulations often are responses to experiences generated within the system and change over time, coevolving with the activities they are meant to regulate (Dopfer and Potts Reference Dopfer and Potts2008).

Routines are at the centre of the evolutionary approach to decision making. Earl (Reference Earl2023, p. 4) reviews the approaches to decision making in old and new behavioural economics as well as evolutionary economics and suggests, ‘The evolving sets of rules, heuristics, principles and routines that decision-makers use to deal with the challenges of everyday life may be genetically inherited, personally created or outsourced/absorbed from social networks, society in a wider sense and market institutions.’ These sets of decision-making rules are the outcome of a process where, ‘Economic evolution entails the creation of new rules and a competitive selection process whereby the relative populations of different rules (or sets of rules) change, with associated changes in the connective architecture of the economic system and of its subsystems’ (Earl Reference Earl2023, p. 3). This conception of decision making underpins much of the evolutionary approach to consumer demand, which is discussed in the next section.

2.2 Consumer Demand

Early contributions to evolutionary analysis of consumer demand attack neoclassical assumptions of optimising behaviour, especially as applied to new goods. Metcalfe (Reference Metcalfe2001) argues optimisation is neither necessary nor useful in analysing consumer behaviour, suggesting the effects of changes in price and incomes can be well handled by focussing on the time and budget constraints facing consumers in a way that also allows analysis of the demand for new goods. Witt (Reference Witt2001) notes the role of the evolution of consumers wants in explaining the avoidance of satiation despite the enormous growth of personal consumption under modern capitalism.

Despite this early work, Nelson and Consoli (Reference Nelson and Consoli2010) lament the absence of a modern evolutionary theory of household consumption behaviour. They then sketch the outlines of such an evolutionary theory, starting with the requirement that individual rationality is bounded. Households are viewed as engaging in activities to satisfy their wants subject to constraints of income and time, with the degree of success depending on the household’s ability to effectively coordinate its activities. Learning is important to improving this ability, meaning ‘consumption decisions need to be recognized as largely a matter of routine plus marginal changes in routine, except when the household is facing circumstances that are significantly new to it’ (Nelson and Consoli Reference Nelson and Consoli2010, p. 678). Learning is particularly important in the context of new goods and services, ‘the response of households to the availability of new goods and services may involve a significant reorientation of their targets and goals, which in many cases only can be accomplished in the course of learning to do new things’ (Nelson and Consoli Reference Nelson and Consoli2010, p. 681–682).

Chai and Babutsidze (Reference Chai, Babutsidze, Dopfer, Nelson, Potts and Pyka2024) review the substantial body of work that has been done since Nelson and Consoli (Reference Nelson and Consoli2010) to fill gaps in evolutionary consumer theory. Notable advances have occurred, including in the modelling of bounded rationality, understanding the impact of innovation in consumer goods on escaping the satiation trap implied by Engel Curves relating consumption expenditure to income, and the role of social networks in the diffusion of new goods. Chai and Babutsidze (Reference Chai, Babutsidze, Dopfer, Nelson, Potts and Pyka2024, p. 270) conclude, ‘This body of work has contributed to developing a more realistic understanding of i) consumer decision-making, ii) the process of preference formation, iii) the role of consumers in the innovation process, iv) the diffusion of innovations among heterogenous consumers and v) the co-evolution of demand and supply.’

For long-period analysis, treating consumer preferences as exogenous, as in neoclassical theory, is not useful. Consumer preferences are impacted by structural change in the economy as well changes in demographics, culture, and politics. For example, current real wages affect housing status, family formation, and population growth, which in turn affect the future age structure of the population, labour force, output, and consumer demand.Footnote ³

Consumer demand needs to be considered as part of the coevolutionary process involving the economy and all other aspects of society, including the process of change in economic, cultural, educational, and political institutions (Dopfer and Potts Reference Dopfer and Potts2008, Almudi et al. Reference Almudi and Fatas-Villafranca2021). Potts (Reference Potts2017) points to the coevolution of institutions and consumer demand in explaining why Keynes (Reference Keynes1930) was so wildly wrong in his prediction regarding the expansion in leisure that would accompany growth in productive capacity over time. Potts argues Keynes ignored innovation under capitalism is endogenous and is oriented towards generating profit, which means a bias towards encouraging production and consumption of new goods over pure leisure. Expanding production under capitalism cannot solve the economic problem of scarcity because creating markets (and the preferences underlying demand) attracts the attention of entrepreneurs.

Time is required for novel products to build up a substantial market. Earl (Reference Earl2022, Chapter 11.7) refers to the uptake of novel products as following a meso trajectory, with the number of users of a novel product increasing over time along an S-shaped curve. In the origination phase, the novel product attracts a niche market of pioneering buyers who have a particular interest in the distinguishing features of the novel product, and who learn about the product through their social networks or other specialist sources. Initially, the product is often expensive relative to available alternatives, and to its own future price, due to high start-up production cost.

Earl (Reference Earl2022, Chapter 11.9) cites several barriers to the rapid adoption of new technology, including the high initial cost to the consumer, the uncertainty regarding the success of competing standards for the new technology (as in Betamax versus VHS for video players), and limited availability of complementary products (as in the limited proportion of all music available on CDs). As prices of the new technology fall, technological standards are clarified, and availability of complements expands as more consumers are attracted. Eventually, new lifestyles evolve incorporating the novel product, as has been the case of with household appliances, including washing machines, dishwashers, and air conditioners. Suburban lifestyles based on the automobile with shopping centres and long distances to work provide an excellent example of the coevolution of novel products and institutions.

When novel products reach the maturity phase of the meso trajectory, they have been acquired by very large proportions of the target population. In the case of durable products, new sales depend on replacement demand. Marketing efforts switch to convincing consumers to replace their current model with something new and better, while for nondurables the effort is on increasing the quantity consumed by each consumer. Saturation of the market limits the growth of demand, with the novel product having reached the top of the S-shaped diffusion curve.

Where does the discussion of this section leave an evolutionary theory of consumer demand? The mainstream assumption of unlimited calculation ability and perfect information are clearly rejected. The notion that buyers always can identify and transact at the best available offer is unviable. Their purchases are subject to the incomplete distribution of information on prices and product availability, introducing a stochastic element into the amount bought at any price. Yet, as Nelson and Consoli (Reference Nelson and Consoli2010, p. 679) point out, ‘the “demand curves” described here are capable of doing many of the same jobs as the demand curves depicted and rationalized in neoclassical theory.’

Differences between the ‘demand curves’ of evolutionary theory and those of neoclassical theory are generally a matter of degree, at least for goods and services for which consumers have established routines. Constraints on information and decision-making ability encourage reliance on habits and routines that lead to inertia in consumer behaviour. Consequences include limited response to changes in income and prices as compared to optimising behaviour, especially in the very short run. Hence, the short-run price elasticity of demand for mature consumer goods and services is likely to be low. Likewise, the short-run marginal propensity to consume out of income is expected to be below the corresponding average propensity. Over time, greater adjustment occurs, better understood as a lagged response than a higher long-run elasticity, as path dependency means adjustments continue even if the price or income change is reversed.

Foster (Reference Foster2021) uses the evolutionary approach outlined earlier to develop a model of aggregate consumption building on the Keynesian approach of Dusenberry (Reference Duesenberry1949). Aggregate consumption has two components. The first component ‘is determined by a precommitment to a particular set of interconnected behavioural rules, ranging from meso-rules that are broad and cultural, down to personal commitments to individual routines and habits that are influenced by targeted advertising and marketing’ (Foster Reference Foster2021, p. 783). This component accounts for the bulk of consumption expenditure and is heavily path dependent, so current aggregate consumption depends heavily on past consumption. The second component consists of purchases of novel goods and services and follows a diffusion process as awareness and interest in these novelties spread through networks. Diffusion builds the impact of expanding connections into aggregate consumption through networks of individuals, with limits set by the size of the relevant population.

Foster estimates parameters of the model with data from the US economy from 1972 to 2018. The results are consistent with the hypothesised relationship and suggest a gradually waning impact of consumerism, which is the dominant meso-rule stimulating consumer demand for novelties since the middle of the twentieth century. As a result, the impact of fiscal policy on growth is weaker than in earlier decades and greater expansions of money or budget-financed credit are required to keep unemployment low.

2.3 Conclusions

The mainstream characterisation of the representative consumer maximising utility with exogenously determined preferences is unhelpful, even destructive to understanding economic evolution. Individuals are heterogeneous, with differentiated knowledge, cultures, and experiences. They have imperfect information and limited cognition, which leads to reliance on habits and routines in decision-making. They are also occasionally creative, especially in dealing with disappointed expectations, novel circumstances, and new products.

Consumer demand reflects the use of routines and habits to deal with imperfect information and limited cognition in a changing and uncertain environment. Demand takes time to adjust to changes in price or income, while demand for new products depends on diffusion of interest through networks of individuals. Evolutionary analysis resolves the satiation problem, as demand coevolves with the supply of novel products in an unending process moulded by institutional frameworks and social networks.

3.1 Firm Size, Scope, and Organisation

A better understanding of the variety of behaviour across firms can be achieved by distinguishing categories of firms according to size, scope, and organisation. Firm size is important as only larger firms can internally reap full advantages of specialisation from the division of labour. However, specialisation within the firm poses challenges of coordination and resistance to innovation. These challenges increase with the scope of firm activities, but increased scope offers opportunities for diversification and synergy, especially in relation to innovation through new products, processes, and markets.

3.2 Orientations of Firms

Evolutionary economics is history friendly. Path dependency, historical events, and individual personalities feature prominently in the accounts of the development of individual firms and industries (Chandler Reference Chandler1962 and Reference Chandler1977, Malerba et al. Reference Malerba, Nelson, Orsenigo and Winter1999, Dopfer Reference Dopfer, Foster and Metcalfe2001). The observed variety of firms in the economy reflects the importance of their separate experiences. It also reflects the different ways in which firms interact with their environment. In this section, four orientations of firms are discussed, survival, profit, growth, and innovation. A firm may have multiple orientations, which are often complementary as noted at points in the discussion next.

3.3 Pricing Behaviour

Mainstream economics portrays competition as a structural characteristic of markets determined by the number and relative size of buyers and sellers. This is highly misleading, ignoring the roles of technology, institutions, and history in determining patterns of competition. Small firms often operate in a niche market with limited competition, such as the local bakery or a highly specialised professional service provider. Also, very large firms operating in markets for standardised commodities often have very limited control over the current market price of their product, with examples including firms buying and selling commodities like wheat or coal on world markets.

All firms in an evolving economy regardless of size, scope, and organisation face an environment characterised by imperfect information and uncertainty. They also generally operate with a limited number of direct competitors. Crucially, the historical and institutional characteristics of the market are critical in creating the competitive conditions faced by a firm, more so than the number or size distribution of firms emphasised in mainstream economics. Nonetheless, the distinction between price-taker and price-maker firms that features in mainstream theory of perfect and imperfect competition can be usefully adapted to evolutionary price theory.

3.3.2 Price-Maker Firms

Price-maker firms use rules or routines to set a fixed price for each of their products and offer all their available supply at those prices. Prices set by these firms are indirectly impacted by influences of buyer demand for the products or the pricing behaviour of firms selling substitute products. The indirect impact on rules and routines used in pricing is explained in the remainder of this section.Footnote ⁶

Bloch (Reference Bloch2018b) suggests using post-Keynesian pricing rules in developing micro-level analysis for an evolutionary theory of price determination. Post-Keynesian pricing rules are designed to capture the behavioural routines used by firms possessing market power, especially firms that dominate manufacturing in the modern economy. Large and complex enterprises need behavioural routines that can be applied across the organisation and monitored centrally. Importantly, post-Keynesian pricing rules share the presumptions of evolutionary analysis, namely imperfect information, distributed knowledge, open systems, and development from within.

A general characterisation of post-Keynesian pricing rules is that firms set price equal to a measure of normal unit cost multiplied by the firm’s desired price-cost ratio (Lee Reference Lee1999, Bloch Reference Bloch, Courvisanos, Doughney and Millmow2016b). Normal unit cost is the level of cost when the firm operates at its expected level of capacity utilisation, which is generally well below maximum possible production. Excess capacity provides a buffer for supplying unexpected increases in demand and a strategic deterrent against existing and potential rivals. In some post-Keynesian pricing rules, the measure of unit cost covers the normal full cost of production, including overheads and an allowance for the depreciation of fixed capital. In other variants only labour, raw materials, and intermediate inputs used in current production are included, providing a measure of normal average variable cost. A feature common to all variants is that the measure of normal unit cost doesn’t change as output fluctuates in the short period.

Various explanations are given regarding how the price-cost ratio desired by firms is determined. The full-cost pricing rule of Hall and Hitch (Reference Hall and Hitch1939) emphasises dealing with uncertainty and maintaining long-run sustainability. Monopoly power features prominently in the theoretical work of Kalecki (Reference Kalecki1971) on mark-up pricing, with the concept of monopoly power extending beyond the market structure notion of neoclassical economics (Kreisler Reference Kriesler1987). Requirements for internal financing of firm investment in extra capacity feature in the work of Eichner (Reference Eichner1976), Harcourt and Kenyon (Reference Harcourt and Kenyon1976), and Wood (Reference Wood1975). Steindl (Reference Steindl1976) and Sylos-Labini (Reference Sylos-Labini1962) emphasise the rate of growth of industry demand as well as conditions of entry and exit for the industry in which the firm operates.

A general form of post-Keynesian pricing rules has normal unit cost and the desired price-cost ratio as the proximate determinants of the administered price set by price-maker firms in the short period. The price for the k^th firm operating in the i^th industry at time t, $p_{k,i,t}$ , is given by the firm’s normal unit cost, $u_{k,i,t}$ , multiplied by the desired price-cost ratio, ${\text{π}}_{k,i,t}$ ,

(3.1)$p_{k,i,t}=u_{k,i,t}{\text{* }\text{π}}_{k,i,t}$

Fluctuations in output due to rising or falling demand have no immediate impact on the desired price-cost ratio or normal unit cost, so the short-period price is unaffected by demand fluctuations. In contrast, changes in prices for intermediate inputs and labour, unless temporary and expected to be reversed, lead to changes in normal unit cost and proportional changes in product price to maintain the desired price-cost ratio.

Key questions in applying post-Keynesian pricing rules to an evolutionary context are what determine the values of normal unit cost and the desired price-cost ratio both at a point in time and in terms of movement over time. Bloch (Reference Bloch, Courvisanos, Doughney and Millmow2016b) addresses these questions in the context of providing a Schumpeterian twist to post-Keynesian price theory. Key issues identified there are whether the costs of overhead and depreciation are included in the unit cost measure or in the desired price-cost ratio, the role of market structure in determination of the desired price-cost ratio, the impact of creative destruction on the dynamics of industry-average values of normal unit cost and desired price-cost ratio, and how obsolescence and R&D costs are handled. These issues are discussed at various points in the following chapters.

3.3.3 Price Leadership

Price-taker and price-maker firms can coexist in the same market. Indeed, models of price leadership presume such coexistence. In the simplest case of price leadership, firms produce virtually identical products and prices for all firms are equal. A single dominant firm, the leader, is the price maker and other firms, followers, act as price takers. Heterogeneity in normal unit cost across firms is accommodated through compensating differentials in price-cost ratios. The price-cost ratio for each of the price-follower firms is determined by its own normal unit cost and the price set by the leading firm, firm d,

${\text{π}}_{k,i,t}=p_{d,i,t}/u_{k,i,t},\text{for}k\ne d\text{and}{p}_{d,i,t}=u_{d,i,t}{\text{* π}}_{d,i,t}$(3.2)

Price leadership is an extreme case of interdependence of pricing rules across firms producing similar products. Kalecki (Reference Kalecki1971, Chapter 5) proposes a pricing rule with a firm’s price as a linear function of its own normal unit cost and the average price of related products,

$p_{k,i,t}=(m_{k,i,t}*u_{k,i,t})+n_{k,i,t}*({{\sum }_{k}s}_{k,i,t}*p_{k,i,t}),m_{k,i,t}\ge 0,n_{k,i,t}\le 1$(3.3)

The last term on the right-hand-side of Equation (3.3) is the weighted average price of all firms producing similar products, with $s_{k,i,t}$ being the share of industry sales accounted for by the k^th firm. A greater weight on own cost suggests greater independence in pricing and a greater weight on average price suggests greater interdependence. In the limiting case of price leadership, $m_{jk,i,t}$ > 1 and $n_{k,i,t}$ = 0 for the price leader, while $m_{k,i,t}$ = 0 and $n_{k,i,t}$ = 1 for price followers (Asimakopoulos Reference Asimakopoulos1975, Bloch Reference Bloch1990).

The firm’s price increases with both the $m_{k,i,t}$ and $n_{k,i,t}$ coefficients, given levels of its own unit cost and the average price of similar products. These pricing coefficients can vary over firms, with the price leadership case being an extreme example. The average price across all firms is given by,

${\overline{p}}_{i,t}=\left[{\overline{m}}_{i,t}/\left(1-{\overline{n}}_{i,t}\right)\right]{\overline{u}}_{i,t}$(3.4)

where a bar over a variable or coefficient indicates it is an appropriately weighted average over the group of products as with average price in the last term of (3.3). Kalecki (Reference Kalecki1971, Chapter 5) refers to the term $\left[{\overline{m}}_{i,t}/\left(1-{\overline{n}}_{i,t}\right)\right]$ as the degree of monopoly for the industry consisting of the group of producers. The degree of monopoly is influenced by a range of factors, including industry concentration, the degree of sales promotion, overhead costs as a proportion of total costs, and the strength of trade unions.Footnote ⁷

The expressions in (3.3) and (3.4) are useful in evolutionary price theory because they connect individual firm price setting to the firm’s own normal unit cost and the average price in the industry. Innovation by a single firm impacts its cost and price, which then flows through to the prices of other firms in the industry. Average price for the industry is determined by the averages of pricing coefficients and unit cost across all the firms in the industry, which means that average price changes with the distribution of market shares as well as changes in pricing coefficients and normal unit cost at individual firms. Changing structure of market shares due to differential firm growth is at the centre of analysing price dynamics for markets disrupted by innovation in Chapter 5.

3.3.4 Pricing of Novel Products

Novel products are central to the evolutionary process. Distinctive new products feature prominently in consumer expenditures over the modern era, products such as computers and mobile phones most recently, adding to televisions and air conditioners for the previous generation, automobiles and radios for the generation before, and the even earlier mass production of textiles, clothing, and footwear. Such products have been essential in maintaining consumption as a share of household income despite rising real incomes noted by Foster (Reference Foster2021) as discussed in Chapter 2.

Novel producer products have been essential for transforming production processes. Mechanised assembly lines, electrification, computers, and robots are among the array of novel products raising labour productivity and lowering manufacturing costs. Railroads, trucks, planes, and fossil-fuel powered ships have connected manufacturers to cheaper sources of raw materials and intermediate products. In primary production, costs have been dramatically reduced and output increased using novel products such as tractors, synthetic fertilisers, seismic mapping, dredge lines, computers, and drones.

Pricing of novel products is complicated. The producer faces great uncertainty regarding buyer reaction and the cost of production over the full meso trajectory of origination, adoption and retention for the product as discussed in Chapter 2. Mainstream economics ignores the issue. Novelty is inconsistent with the certainty presumed in perfectly competitive equilibrium.

Pricing analysis discussed in previous sections is compatible with novel products, but a nuanced and multi-phase application is required. Nuance is required in combining elements of the analysis of price takers, price makers, and price leadership. Further, separate analyses are required for each phase of the meso trajectory of origination, adoption, and retention of the novel product.

Initially, the producer of the novel product is a price maker without close competitors, but its circumstances differ from those of an established seller who is a monopolist. In particular, the producer of the novel product can take the prices of its distant competitors as given because it looms small as a threat when its product is first introduced. In this sense, the firm starts in a situation with somewhat like a follower firm under conditions of price leadership.

In its situation as a price follower, the prices of established products provide benchmarks for the producer of the novel product. For consumer goods, the buyer needs to fit the novel product into the established pattern of expenditure. Comparisons to existing products that satisfy related wants are inevitable, with price as well as product characteristics evaluated in terms of value for money. For producer goods, the novel product is attractive when the downstream producer can lower its production cost. The price of currently purchased producer goods determines the comparison cost.

Pricing novel products using an adapted version of Kalecki’s model of price-setting equation, (3.3), includes a benchmarking role for prices of established firms. The price of the novel product of an entrepreneurial firm, e, that starts a new industry, n, is given by,

$p_{e,n,t}=(m_{e,n,t}*u_{e,n,t})+n_{e,n,t}*{\overline{p}}_{c,t},\text{where }{n}_{e,n,t}>0\text{ and }{\overline{p}}_{c,t}=({{\sum }_i{\sum }_{k}s}_{k,i,t}*p_{k,i,t})$(3.5)

Weights used in calculating the average price of competing products, ${\overline{p}}_{c,t}$ , are shares of prospective customers for the novel product who are currently buying products of other firms scattered across established industries. Each competing price is scaled to a quantity equivalent to a unit of the novel product. The value of the parameter, $n_{e,n,t}$ , is determined based on the competing needs of the entrepreneur for a high price to generate profits to finance growth in productive capacity and for a low price to attract customers from competitors.

In the origination phase of the meso trajectory of a novel product is viable only if its price exceeds it unit cost. If $m_{e,n,t}$ is set equal to one and both sides of (3.5) are divided by unit direct cost of the novel product, the resulting equation for determining the product’s price-cost ratio is,

$p_{e,n,t}/u_{e,n,t}=1+n_{e,n,t}*\left({\overline{p}}_{c,t}/u_{e,n,t}\right)$(3.6)

The novel product is a value proposition to customers of competing products when,

${\overline{p}}_{c,t}/p_{e,n,t}=\left({\overline{p}}_{c,t}/u_{e,n,t}\right)/(p_{e,n,t}/u_{e,n,t})>1$(3.7)

This occurs only if ${\overline{p}}_{c,t}/p_{e,n,t}<{\overline{p}}_{c,t}/u_{e,n,t}$ , which only occurs if the unit direct cost for the novel product below the average price of competing products and the value of $n_{e,n,t}$ is substantially less than one.

Keeping the price-cost ratio high during the origination phase generates profits to provide internal finance for investment in capacity expansion and can help in attracting external finance. Perhaps, even leading to acquisition by venture capitalists or an established firm with cash flow to invest in rapid expansion of production capacity. Pioneering customers provide a market for limited output in the origination phase, even when the price of the novel product doesn’t provide much better value than alternatives.Footnote ⁸

During the adoption phase of the meso trajectory of the novel product, a key consideration is the threat of entry from imitators. The benchmark for prices of competing products shifts from the prices of established distant competitors to prices likely to be charged by imitators. Sylos-Labini (Reference Sylos-Labini1962) suggests the entry-limiting price is equal to the unit cost of a small-scale potential entrant. Prices above this level attract at least some entry. A trade-off exists between ceding market share to imitators and generating internal finance for expansion of production capacity, as discussed in Section 3.3.3.

In the retention phase of the meso trajectory for the novel product, there are imitators operating in the industry. Here, price dynamics depend on the interaction between the innovating firm and its imitators, who have unit direct cost that may be below, equal, or above those of the innovator. Price determination for this situation is a special case of price theory for industries disrupted by innovation, which is presented in Chapter 5.

Learning and specialisation of labour in production occurring during the adoption and retention phases of the meso trajectory for novel products contribute to sharp decreases in the cost of production. The dynamics of differential firm growth as discussed in Chapter 5 ensure that cost decreases pass into price decreases, with stable or falling price-cost ratios. Thus, a sharp downward path for prices of novel products relative to those of established goods and services is commonly observed. Witness the experience of prices for computers, mobile phones, electric vehicles, and internet services over recent decades.

3.4 Summary

Firms are heterogeneous, even across firms producing similar products. They face an uncertain future and operate with limited knowledge. Their behaviour is generally regulated by rules and routines, particularly for large firms and mega-firms, who depend on such rules and routines to coordinate the activities of many individuals with specialised knowledge and skills. These rules and routines vary across firms, contributing to the variety of firm behaviour.

Several dimensions of differences among firms are examined in this chapter. First, differences in firm size, scope, and organisation are addressed, with emphasis on differences between small, large, and mega-firms. Second, the orientations of firms are distinguished in terms of survival, profit, growth, and innovation. Third, price-maker firms are distinguished from price-taker firms, both in terms of market conditions facing the firms and the behaviours they adopt in response. Fourth, price leadership is discussed as involving both price-taker behaviour (for the followers) and price-maker activity (for the leader). Finally, the pricing analysis is applied to determining price for a novel product with pricing behaviour changing between the origination, adoption, and retention phases of the meso trajectory for the novel product.

4.1 Coordinating Price

Schumpeter (Reference Schumpeter1939) refers to a theoretical norm for price in discussing the prices that occur when innovations are fully absorbed in the economy. Marshall, his followers, and Sraffians use related concepts of normal price and reproduction price, respectively, to refer to the prices that occur in the long period, when external disturbing influences have dissipated (Bloch Reference Bloch2022). Forces determining the long-period price in these approaches are conceptually distinct from each other (Bloch Reference Bloch2020).

For reasons explained in remaining chapters, especially Chapter 7, Schumpeter’s concept of a theoretical norm for prices is rejected as being logically incoherent and unhelpful in evolutionary price theory. Instead, coordinating price is suggested as an appropriate theoretical concept for price. The coordinating price is a price consistent with orderly exchanges between buyers and sellers in a market in the short period, but subject to change over time as the market develops from within. No concept of a disturbance-free or long-period price is offered for evolutionary price theory as disturbance is always present or emergent in the evolving economy. Evolutionary price theory is a theory of prices in motion.

The remainder of this chapter discusses the role of price information in coordinating buyer and seller activities in an evolving economy. The concept of market order replaces the concept of market equilibrium in mainstream price theory. First discussed is price determination in orderly markets with large numbers of buyers and sellers, adapting the concepts of supply and demand to an evolutionary context. Following is a discussion of using administered prices to create order in markets dominated by a small number of sellers or buyers. With either market-determined prices or administered prices, coordinating price is the theoretical price concept for market order.

Buyers and sellers follow fixed routines of behaviour of the types discussed in the last two chapters. Constancy is also assumed for the structure of production, skills, resources, and equipment, so the analysis is along the lines of Marshall’s (Reference Marshall1920) analysis of the short period. Analysis in the next chapter addresses the process of adjustment of routines and the structure of production over time, roughly corresponding to Marshall’s concept of the long-period adjustment. The coordinating price follows a trajectory from one short period to the next during the adjustment process.

Throughout this chapter and the next, the analysis focusses on individual markets. The individual market is a sub-system of the economy. Order in a market is subject to disturbance from within if there is incomplete adjustment to impact of innovations, as is discussed in the next chapter. Order in a market is also subject to external disturbance as other markets adjust to the impact of innovations. The analysis of individual markets is extended to the interaction across markets, and to the aggregate economy, in Part III.

4.2 Spontaneous Market Order

The market economy is a system that connects groups of heterogeneous agents and organisations, individuals, households, firms, regulatory agencies, governments and others, who have disparate knowledge, skills, and resources. Agency implies objectives or goals and the means to achieve these goals. Connections involve the exchange of information between agents that result in the distributed and simultaneous attainment of goals. Prices are essential components of this information providing terms on which transactions occur in the pursuit of these goals.

For analytical purposes the market economy can be thought of as a collection of market sub-systems. In a complex economy as a system, interactions are numerous within sub-systems relative to interactions between sub-systems. Thus, a single system can be considered as if it were isolated or considered as connected to other systems to form a hierarchical or layered interaction (Arthur Reference Arthur2015).

Buyers and sellers, operating under their respective routines as discussed in Chapters 2 and 3, generally interact in markets under conditions of imperfect information. Each buyer is aware of a subset of information about the prices, characteristics, and availability of goods or services being offered by sellers, while sellers are generally imperfectly informed of the extent of interest from buyers. Not every buyer or every seller need end with the best possible deal or even with a transaction. Yet, there can still be spontaneous order among buyers and sellers in the market.

Nelson (Reference Nelson2013) rejects assumptions of optimisation and equilibrium from mainstream analysis of markets as inconsistent with continuing endogenous structural change in an evolving economy. Nonetheless, most markets in an evolving economy are far from chaotic at any point in time. Nelson (Reference Nelson2013, p. 29) suggests applying the concept of market order, providing the definition, ‘An orderly market is characterized by a set of routines established over time that when employed by potential buyers and potential sellers are tuned to each other and generally result in transactions that are satisfactory for most parties on both sides of the market.’ He then suggests using supply and demand analysis based on behavioural routines, rather than optimisation, to analyse price determination in orderly markets with large numbers of both buyers and sellers. Order is created as the routines of buyers lead to increased purchases at lower prices, while supplier routines result in offering to sell more at higher prices.Footnote ¹⁰

An orderly market need not have a single price, even for a homogeneous product. Imperfect information about prices, even about the identity of potential buyers and sellers, means transactions can occur at differing prices. Likewise, there can be missed transactions when willing buyers and sellers aren’t connecting. Neither price nor quantity need be fully determined by the willingness of buyers and sellers to engage in transactions.

With imperfect price information and imperfect connections between buyer and sellers, supply and demand are not interpretable as the single valued functions used in mainstream economics. Rather, they represent sets of buyers and sellers willing to transact at various prices. In Figure 1, the demand set boundary, D, shows the maximum price (willingness to pay) that can be achieved for various volumes of a product offered to buyers, while the supply set boundary, S, shows the minimum price required to elicit various volumes of product from sellers. The intersection of the boundaries D and S gives a theoretical maximum quantity, $Q_C^{*}$ if all possible transactions occur at the coordinating price, $p_C^{*}$ .

Figure 1 Order with a market-determined price

If all potential buyers and sellers are connected and if all transactions occur at the price where the boundaries of the demand and supply sets intersect in Figure 1, $p_C^{*}$ , we have the special case of perfect market order,

$Q_c={\sum }_{\text{i}}^{\text{n}}\left(q_i\right)=Q_C^{*}\text{and}{p}_c={\sum }_{\text{i}}^{\text{n}}\left(q_i{p}_i\right)/Q_c=p_C^{*},\text{as}{p}_i=p_C^{*}\text{ for all }i\text{in}n,$(4.1)

where $q_i$ and $p_i$ are the quantity and price for the i^th transaction out of n. Importantly, Equation (4.1) doesn’t imply the market is in equilibrium. Equilibrium implies the absence of any possibility of change from within. Perfect market order occurs when all possible buyers and sellers are connected and all trades occur at the coordinating price, $p_C^{*}$ . The possibility of endogenous change remains. Also, there is no implied tendency for perfect market order to be restored if trades occur away from the coordinating price.

The outcome of perfect market order is extremely unlikely in a competitive market with imperfect information and incomplete connections between buyers and sellers. Individual voluntary transactions can occur anywhere within the set boundaries in Figure 1. Not all possible transactions need occur, and many transactions will occur at prices above or below the coordinating price.

Nelson’s (Reference Nelson2013, p. 29) definition of market order doesn’t require perfect order in terms of Equation (4.1); rather, he requires only, ‘transactions that are satisfactory for most parties on both sides of the market.’ In an orderly market $Q_c$ won’t fall too far short of $Q_C^{*}$ and $p_c$ won’t diverge too far from $p_C^{*}$ . Additionally, to avoid widespread buyer or seller regret, the variance of $p_i$ should not be large. Thus, conditions for market order are of the form,

$Q_c=Q_C^{*}-\alpha ,p_c\approx p_C^{*},\text{α}<\text{φ and Var}\left({\mathrm{ p}}_i^{*}\right)={{\sum }_{\text{1}}^{\text{m}}\left(p_i-p_c^{*}\right)}^2/m<\theta ,$(4.2)

where $\text{φ}$ and $\theta$ are upper limits on missed voluntary transactions and variance of transaction prices, respectively. These requirements for a market to be considered orderly lack quantitative precision, at least at this stage of development of the theory of market order.

Institutional arrangements can be expected to affect the extent to which quantity traded in a market falls short of $Q_C^{*}$ , the degree to which average price diverges from $p_C^{*}$ , and the size of the variance of $p_i$ . Many types of institutions exist, from local farmer’s markets to organised commodity exchanges, such as the London Metals Exchange and the Chicago Board of Trade. Institutions have evolved over time from medieval town markets to internet platforms, such as eBay. Changes in transport, communications, and information technology have all had an impact. All of this is ignored in mainstream analysis through assuming perfect information and zero transaction costs.

Bid and ask prices on organised commodity exchanges provide information on the willingness to trade of buyers and sellers. However, such bid and ask prices can’t be directly interpreted as indicating boundaries to the demand and supply sets for the commodity being traded. A buyer certainly won’t bid more than their willingness to pay, but they may bid less hoping to secure supply at a lower price. Likewise for supplier offer prices. The variability of transaction prices over any trading period makes such strategies plausible if not always successful.

Data on transactions from organised commodity markets include the activities of traders who are neither suppliers nor users of the commodity, but who seek to profit from buying cheap and selling dear. If they are successful, the activities of these traders tend to stabilise prices, helping to achieve market order. Yet, there are occasions when price volatility on organised commodity markets exceeds bounds consistent with market order, even occasions when market rules on excessive price movements lead to trade being halted with buyers and sellers unable to complete any transactions. More generally, market order, even with the loose bounds of Equation (4.2), is not guaranteed by competition among large numbers of buyers and sellers.

Heterogeneity of buyers and sellers is fully compatible with market order. Indeed, variation in bids across buyers provides an explanation for the downward slope of the demand set boundary in Figure 1, as variation in offers across suppliers explains the upward slope of the supply set boundary. For example, market demand may consist of buyers who only purchase a single unit at varying maximum prices, while market supply may consist of sellers who offer a fixed number of units each at varying minimum prices that are above their average variable unit cost of production. Mainstream assumptions of declining marginal utility for each word buyer and increasing marginal cost for word seller are unnecessary.

Evolutionary change is also fully compatible with market order. Learning by buyers as discussed in Chapter 2 leads to changes in the demand set boundary in Figure 1, while growth and innovation by firms as discussed in Chapter 3 lead to changes in the supply set boundary. Whether these changes in set boundaries flow through into equal or proportionate changes in the quantities traded or the average price of the product depends on institutional structure, on information flow and connections between buyers and sellers in the market. Still, in an orderly market, changes in $Q_C^{*}$ and $p_C^{*}$ can be expected to flow through roughly in proportion to changes in the values of $Q_c$ and $p_c$ .Footnote ¹¹

Changes due to external shocks move the boundaries of the demand and supply sets in Figure 1, much as do the changes arising from evolution when consumers are learning, or when firms are growing and innovating. The analysis of market order presented here encompasses both evolutionary change from within and the impact of external shocks that are emphasised in mainstream analysis. Mainstream analysis of equilibrium is restricted to the impact of external shocks because evolutionary change from within is inconsistent with equilibrium.

Econometric practices for using series of market quantities and prices as observations of underlying theoretical values tend to treat observations of prices symmetrically with observations of quantities. This treatment is not correct for orderly markets, according to the analysis leading to the conditions in Equation (4.2). While the average value of price for all transactions in the market is an unbiased approximation of the coordinating price, $p_C^{*}$ , the total quantity traded is a downward biased approximation of the quantity traded under perfect market order, $Q_C^{*}$ . Estimation methods, such as ordinary least squares, that assume a symmetric disturbance of disturbances around a theoretical value are valid for estimating relationships with price as a stochastic variable. However, when quantity is the stochastic variable, a method that takes account of the one-sided distribution of the disturbances, such as stochastic frontier analysis, is required.

Nelson (Reference Nelson2013, p. 33) is certainly on the right track in suggesting, ‘the basic arguments in price theory about how changes in demand and supply affect prevailing prices and quantities hold up under the theory of market order I have been developing, but without the encumbering baggage about the characteristics of market equilibrium.’ He adds cautions about needing to interpret supply and demand curves as heuristic simplifications, about not all potential transactions occurring, and about the importance of institutional details. Each of these cautions is reflected in the discussion earlier.

Importantly, Nelson (Reference Nelson2013, p. 38) points out that the tools of price theory based on market order are useful for short-run analysis, but ‘Understanding of long run economic change requires a very different mode of analysis, which is what modern evolutionary economics is mostly about’. Adjustment of prices and quantities to major innovations is the subject of the next chapter. First, the following section discusses extending the notion of market order to markets without large numbers of sellers or buyers.

4.3 Administered Market Order

Bloch and Metcalfe (Reference Bloch and Metcalfe2018) extend application of the concept of market order to markets with dominant sellers or buyers. They note the use of administrative routines by dominant firms to set prices fits well with the presumptions of evolutionary analysis, providing a reasoned response to the need for internal control to achieve strategic objectives in a complex environment. Firms with radically new products use administered prices to establish market order, even though demand curves in the usual sense don’t exist. Administered prices combine with routines to deal with unexpected fluctuations in demand or supply, such as maintaining inventories or underutilised productive capacity for sellers and accumulating superfluous inventories for buyers. These routines provide an alternative to price adjustments as a means of achieving market order.

As explained in Chapter 3, setting an administered price generally involves multiplying a measure of normal unit cost by a desired price-to-cost ratio exceeding one. With the unit cost measure based on operating at normal output, changes in output need not have immediate effect on prices. In contrast, changes in operating cost due to changes in input prices are fully and quickly passed on to changes in product price. Thus, the pattern of changes in administered prices in response to demand and cost changes is different from the pattern of changes in coordinating price in markets with spontaneous market order, where changes in both demand and cost are partially passed on to price changes.

Maintaining market order with administered prices represents strategic behaviour of firms. For example, with prices set by the seller, market order requires that buyers are generally able to obtain their desired quantities at those prices. Manufacturing firms are usually able to satisfy buyer demand at the set price by operating with substantial inventories or underutilised capacity to accommodate fluctuations in demand or potential supply chain disruptions. The quantities traded are satisfactory to the sellers in the sense that they are consistent with the strategy of survival and growth in the long period.

Quantity determination in an orderly market for a product j with an administered price is depicted in Figure 2. The demand set depicted is of similar form to that in Figure 1, while the supply set shows any quantity is available at the administered price, $p_j$ , up to the capacity of the seller, $q_j^x$ . If all potential buyers can identify the seller and the quantity buyers desire at $p_j$ is below capacity, $q_j<q_j^x$ , then all potential transactions are completed and occur at the administered price, which then constitutes a coordinating price. All buyers willing to pay the administered price are served and the quantity sold is consistent with the firm’s strategy.

Figure 2 Order with an administered price

Multiple firms selling similar products are grouped into an industry for purposes of data collection and for analysing firm interactions, including differential firm growth. Industry quantity sold, $Q_A$ , is given by adding up quantities of individual firms (in physical units) and average price, ${\overline{p}}_A$ , is given by summing values sold across firms and dividing by total quantity sold,

$Q_A={\sum }_{j=1}^{\text{n}}\left(q_j\right)\text{and}{\overline{p}}_A=\left[{\sum }_{j=1}^n\left(p_j*q_j\right)\right]/Q_A$(4.3)

When the same price is charged by all sellers, $p_j={\overline{p}}_A$ , and each seller is serving all potential customers at that price, then, $Q_A=Q_A^{*}$ , where $Q_A^{*}$ is the maximum quantity of sales with voluntary transactions. Perfect market order is achieved with the industry average price as a coordinating price, ${\overline{p}}_A=p_A^{*}$

Market order can be considered in a broader context when firms set differing administered prices. Each seller sets a price and buyers choose amongst the sellers based on the information they have available about sellers and their prices. Missed sales are possible if buyers aren’t aware of sellers whose price is acceptable. Also, buyers need not find the best possible deal. Provided sellers have sufficient capacity to supply all prospective buyers, market order is achieved when,

$Q_A=Q_A^{*}-\text{α },{\overline{p}}_A=p_A^{*}+\beta ,\text{α }<\text{φ},\beta <\text{Φ},andVar\left(p_j\right)={{\sum }_{\text{j}}^{\text{m}}\left(p_j-p_A^{*}\right)}^2/m\le \theta ,$(4.4)

where $\text{φ},\text{Φ}$ and $\theta$ are upper limits on missed voluntary transactions, missed best buys, and variance of transaction prices, respectively, that might be regarded as consistent with market order.

Markets with administered prices are likely to be well ordered, at least in the absence of the type of disruptions discussed in Chapter 5. While each firm faces its own demand set and has its own administered price and supply capacity, the demand sets and determination of administered price are clearly related. As discussed in Chapter 3, price leadership is a common feature of theories of price determination for markets with administered markets. With price leadership and similar products due to competitive imitation, the variance of $p_j$ is likely to be low and the market price and quantity are likely to closely approach their levels associated with perfect market order. Indeed, if all firms set their price equal to that of the market leader, there is no variance in price in the market and the leader’s price is a coordinating price.

For the reasons discussed earlier, markets with administered prices are likely to be more orderly than those where prices are determined by the competitive interactions of large numbers of buyers and sellers.Footnote ¹² Administered prices also are expected to react differently to external shocks than do competitive market prices. As explained in Chapter 3, demand shocks are unlikely to affect administered prices, at least in the short period, while competitive market prices are expected to rise and fall as product demand rises and falls. Thus, ratios in the form of $p_c/{\overline{p}}_A$ can be expected to be procyclical, rising (falling) with positive (negative) demand fluctuations. Further, changes in input prices are likely to flow through to the administered product price roughly in proportion to the input’s share of variable cost, while input prices are expected to pass through only partially into product markets in competitive markets.

Empirically, the administered prices for manufactured goods are generally stable over time relative to the prices of primary commodities used as foodstuffs and raw materials that are largely determined in competitive commodity markets. Building on the work of Kalecki (Reference Kalecki1971), Bloch and Sapsford (Reference Bloch2000) show the world average of primary commodity prices responds much more strongly to short-term changes in world industrial output than does the corresponding world average of manufactured goods prices. Bloch and Sapsford (Reference Bloch, Sapsford, Andersen and Pyka2013) further show the long business cycle associated with Schumpeterian waves of innovation is characterised by procyclical movements in the ratio of primary commodity prices to the average price of manufactured goods.

4.4 Summary

Following on from Nelson (Reference Nelson2013), market order is proposed as a replacement for market equilibrium as a concept for analysing price determination in evolutionary economics. Equilibrium implies no tendency to change from within, which is inconsistent with theory that treats evolution as an endogenous process. Order requires only that most buyers and sellers can execute trades within the bounds of their acceptable prices and desired quantities. Markets with large numbers of buyers and sellers can be orderly in the sense that both the quantity sold and average price approximate those associated with perfect market order, even though individual transaction prices may deviate from that level. The price associated with perfect market order, which is designated the coordinating price, serves as an analytical reference point for evolutionary price theory in parallel to the short-period equilibrium price of mainstream theory.

When prices are set administratively by dominant sellers (buyers), most buyers (sellers) generally are able to obtain their desired quantity of purchases (sales) at the set price. The maintenance of excess capacity or inventories by the sellers or buyers increases the proportion of possible transactions that complete. Price leadership is common in markets with administered prices, which tends to reduce the variation in price across transactions. In the extreme case, there is only one price at which all transactions take place, making this price a coordinating price if there is sufficient capacity for all buyers willing to pay this price. Thus, markets with administered prices tend to be orderly, at least in the short period.

Market order with either competitive or administered prices is supported by stable behavioural routines for both buyers and sellers. Disruptive innovations, such as new products, new production processes, or new market infrastructure, challenge the ability of buyers and sellers to change their routines. Consumers change their budget allocation, firms change their products, production methods and sources of input supply; robots displace workers, web platforms displace brick and mortar outlets, electric engines displace internal combustion engines, etcetera. Disruption of markets and the ensuing process of price adjustment are discussed in the next chapter. System-wide disruption and structural changes at the meso and macro level are then discussed in Part III.

5.1 Reliable Price Information and Financing of Innovations

Market order is under constant threat. Some forms of incremental change, such as gradual growth in demand, are compatible with the maintenance of market order. However, with disruptive innovations, the rules and routines that regulate the behaviour of buyers and sellers no longer lead to expected results. The introduction of substantial novelty into the economy leads to behavioural changes and structural transformation.

Evolving systems are systems open to novelty stimulated unpredictably by their internal operation. Price information is used for market calculation necessary to achieve order. However, as Loasby (Reference Loasby1999, Reference Loasby, Arena and Quere2003) insists, creativity and imagination are equally important as calculative ability in understanding how an economy works. Making and breaking of connections is then reflected in the changing economic structure, which generates further flows of innovation. Prices are essential to this process, through stimulating investment in innovation and then financing investment in expansion of productive capability to diffuse the innovation.

Mainstream economics recognises the coordination aspect of price information in markets, but not the role of price information in stimulating innovations or the impact of innovations on the reliability of price information. While prices underpin the prevailing order, their correlating function, they also guide the formation of conjectures that challenge the prevailing order by setting yardsticks any challenger must meet to displace some existing activity. This is the de-correlating function of prices, to give rise to conjectures about new possibilities and the emergence of novelty. In establishing market order, the price system generates the impetus to challenge that order.

The exchange of information is also a principal reason for continuous and unpredictable changes in the levels of understanding held by agents. Not all new knowledge is reliable (Ziman Reference Ziman1978). A means to test the validity of new conjectures is essential. As Potts (Reference Potts2001, p. 418) notes regarding the dual role of markets, ‘They are spaces where existing knowledge is coordinated and where new knowledge is tested.’ Profitability is a price system test of innovations. Successful innovators use high profits to expand relative to their established rivals in the process of creative destruction.

Static analysis is not appropriate for evolutionary analysis according to Schumpeter ([Reference Schumpeter1934] Reference Schumpeter1961, p. 64), who notes, ‘Development in our sense is a distinct phenomenon, entirely foreign to what may be observed in the circular flow or in the tendency towards equilibrium.’ Yet, the structure of prices at a point in time can still be usefully related to the movement of prices over time. The concepts of coordinating price and market order discussed in the last chapter are concepts related to analysis of the structure of prices at a point in time. In this chapter, the focus is on the movement of prices over time as part of the endogenous process of innovation-driven change under capitalism. In evolutionary price theory, the structure of prices and the movement through time are connected through creative destruction as is explained in Section 5.2.

Schumpeter’s (Reference Schumpeter1947) distinction between adaptive and creative responses is useful in an evolutionary understanding of the way creative agents, with their disparate imaginings of the future, react to the structure of prices. Responding to the existing price structure using established rules, routines, or habits is an adaptive response, which tends to maintain order. A creative response devises a new form of behaviour that is advantageous to the innovator, which tends to destroy order. To paraphrase a favourite Schumpeter example, responding to an increased volume of freight between cities in the early nineteenth century by increasing the number of mail coaches is an adaptive response, whereas building a railway is a creative response.Footnote ¹³

5.2 Differential Firm Growth and the Diffusion of Innovations

Once an innovation is in operation, realised profitability displaces prospective profitability as the criterion used by providers of external financing. The market test of profitability confirms or refutes the projections of entrepreneurs. More directly and importantly, realised profits provide internal financing for expansion by innovating firms. Successful innovators use their extra profits to expand relative to non-innovators. Price dynamics associated with differential firm growth constitute an essential element of evolutionary price theory and this section is devoted to presenting this theory.

Differential firm growth is identified as the mechanism for the diffusion of innovations in modern evolutionary economic analyses following Nelson and Winter (Reference Nelson and Winter1982) and Metcalfe (Reference Metcalfe1998). Output from innovating firms grows relative to that from their established rivals. Contra to the instantaneous and costless diffusion of innovations assumed in mainstream economics, diffusion takes time and involves creative destruction. Creative destruction means the decline and eventual extinction of many established firms, along with obsolescence for capital equipment and redundancy for labour skills.

Metcalfe (Reference Metcalfe1998) analyses industry evolution in terms of changes in firm market shares. His approach is adapted here to examine the dynamics of the market share of a successful innovating firm, denoted by the subscript e. Here, the innovating firm is a new entrant, is externally funded, and is small. The firm then grows faster than the industry average,

$s_{e,i,t}=s_{e,i,0}{\left[\left(1+g_{e,i,t}\right)/\left(1+g_{i,t}\right)\right]}^t,\text{where}\left[\left(1+g_{e,i,t}\right)/\left(1+g_{i,t}\right)\right]>1$(5.1)

where $s_{e,i,t}$ is the share of the innovating firm in industry sales, $g_{e,i,t}$ is the innovating firm’s rate of growth of sales and $g_{i,t}$ is the rate of growth of industry sales. Taking the partial derivative of (5.1) with respect to time gives the rate of change of the share of the entrepreneurial firm,

$\partial s_{e,i,t}/\partial t=s_{e,i,t}*ln\left[\left(1+g_{e,i,t}\right)/\left(1+g_{i,t}\right)\right]\approx s_{e,i,t}*\left(g_{e,i,t}-g_{i,t}\right)>0$(5.2)

Growth in the market share of the innovating firm comes at the expense of non-innovating firms, denoted by the subscript m, with the collective share of non-innovating firms given by,Footnote ¹⁴

(5.3)$s_{m,i,t}={\sum }_{k=1}^n{s}_{k,i,t}=1-s_{e,i,t}$

When growth in the innovating firm’s sales exceeds industry growth, the rate of growth of sales for remaining firms as a group is below the rate of industry growth,

$s_{m,i,t}*g_{m,i,t}=g_{i,t}-s_{e,i,t}\text{*}{g}_{e,i,t}$(5.4)

The extent of this shortfall increases with the innovating firm’s share of sales, becoming negative as the share of the innovating grows,

$g_{m,i,t}=g_{i,t}-\left[s_{e,i,t}/\left(1-s_{e,i,t}\right)\right]\left(g_{e,i,t}-g_{i,t}\right)$(5.5)

Continued growth of the innovating firm requires expansion of production capacity and increased uptake from buyers. Both are linked to pricing but in opposite directions. High prices and profits provide internal financing for expansion as discussed in Section 5.1.2, while low prices stimulate sales to buyers.

The tension between the two informational roles of prices is analysed as a process of differential firm growth. Section 5.2.1 analyses the process in markets consisting of price-taker firms producing a homogenous product with price determined in the market by equating full-capacity production and demand. Section 5.2.2 analyses the process in markets with imperfect competition and administered prices. Section 5.3 then provides a comparison to post-Keynesian theories of pricing, particularly the work of Steindl (Reference Steindl1976).

5.2.1 Market-Determined Prices

Nelson and Winter (Reference Nelson and Winter1982) examine Schumpeterian competition among initially identical firms seeking improved production techniques through R&D expenditure. Each firm has a routine deciding how much, if anything, to spend on R&D, which gives them of a probability of finding a cost-reducing innovation. Successful R&D allows the firm to increase output per unit of input and reduce the cost of each unit produced. Linear production technology is assumed, so that unit operating cost is constant up to capacity and then rises vertically. Firms expand capacity over time based on realised profitability.

The analysis of the impact a cost-reducing innovation is illustrated in Figure 3. All firms start with an operating cost of u per unit and a full cost of c per unit. Combined production capacity is $Q_c$ , which leads to a coordinating price of $p_c$ determined by the interaction of buyers and sellers. If the gap between price and unit cost just covers depreciation and R&D expenditures, price equals full cost per unit of output for all producers. No net investment occurs and coordinating price and quantity remain at $p_c$ and $Q_c$ , respectively.

Figure 3 Competitive price with cost-reducing innovation

The initial impacts of cost-reducing innovation on costs, productive capacity, and price are that a successful innovation reduces the unit operating cost of the innovating firm to $u^{\prime }$ and the unit full cost to $c^{\prime }$ for $q_e$ units of output. Nelson and Winter assume innovations increase output without any change in inputs, which implicitly treats the innovation as increasing productive capacity. Instead, it is assumed here that operating cost per unit of output falls but there is initially no change in productive capacity at the innovating firm, which provides a clearer separation of the impact of cost reduction from that of capacity expansion.

Industry output is initially unaffected by the innovation, so there is initially no change in price following the cost-reducing innovation. The innovating firm is earning extra profits at the prevailing price and, in response, invests in additional productive capacity. Output at other firms is unaffected. If capacity at the innovating firm expands from $q_e$ to $q_e^{\prime }$ , industry output expands to $Q_C^{\prime }$ and the coordinating price falls to $p_C^{\prime }$ .

Even after the fall in price, price remains above unit full cost for the innovating firm, so that firm continues to earn net profit and continues to expand its production capacity. Its extra output contributes to further declines in price. The reduced price is below the unit full cost for non-innovating firms, which means their net profits are negative. They continue operating as long as price remains above unit operating cost, but cease investments in capacity expansion. Thus, the market share of the innovating firm expands along with growth in the market.

In the Nelson and Winter model, each firm maintains the same R&D routine over time, with at least some non-innovating firms spending on innovative or imitative R&D. There is also continued spending on innovative R&D from the innovating firm. In each period, there is a positive probability that such spending results in a successful innovation. These innovations reduce the unit cost of one or more non-innovating firms to $c^{\prime }$ in the case of imitative R&D, or possibly to some level below $c^{\prime }$ in the case of innovative R&D. Continued R&D spending by the innovating firm can also result in lowering unit cost below $c^{\prime }$ .

Nelson and Winter analyse a full cycle of evolutionary process, including generating variety in technology through cost-reducing innovation, reproducing the technology through investment in expanding production capacity, and selection of technology through differential firm growth. Variation in technology is the outcome of a stochastic process, so their analysis relies on computer simulation models to analyse the dynamics of innovations, costs, profits, investments, and market shares.

Metcalfe (Reference Metcalfe2007) narrows the focus of analysis to the reproduction and selection impacts of differential firm growth that are amenable to analytical solutions. The products of all firms are homogenous and are sold at a single price. Unit cost differences in this environment translate inversely proportionally to price-cost ratios,

$({\text{π}}_{k,i,t}/{\text{π}}_{j,i,t})=1/({\text{c}}_{k,i,t}/{\text{ c}}_{j,i,t})$(5.6)

Adapting assumptions from the Nelson and Winter (Reference Nelson and Winter1982) model to the limited scope of Metcalfe’s (Reference Metcalfe2007) analysis, the rate of growth of productive capacity at the k^th firm, $g_{k,i,t}$ , equals its net investment divided by capital in the prior period. Thus, $g_{k,i,t}$ equals the amount by which the firm’s ratio of price to unit full cost, ${\text{π}}_{k,i,t-1}$ , exceeds one multiplied by its ratio of net investment to net profit, ${\theta }_{k,i,t-1}$ , and divided by its ratio of capital stock to full cost, ${\Phi }_{k,i,t-1}$ ,

$g_{k,i,t}=\left({\text{π}}_{k,i,t-1}-1\right)({\theta }_{k,i,t-1}/{\Phi }_{k,i,t-1})$(5.7)

Metcalfe (Reference Metcalfe1998) suggests an analogy between differential firm growth dynamics and replicator dynamics in biology. Fitness in biology is associated with a sub-species increasing its share of the population to which the sub-species belongs. Likewise, fitness of a firm is indicated by an increase in its share of sales among the population of firms selling similar products. An industry consisting of all such firms is a population with a common market selection environment. Metcalfe’s measure of firm fitness is given by the rate of change in its market share, taken to be,

(5.8)$\partial s_{k,i,t}/\partial t=s_{k,i,t}*\left(g_{k,i,t}-g_{i,t}\right),\text{where}{\sum }_{k=1}^n\left(\partial s_{k,i,t}/\partial t\right)=0$

A positive value for the change in share in (5.8) indicates fitness. The innovating firm whose share growth equation is given in (5.2) exhibits fitness in this sense. Notably, fitness is a relative measure, indicated by the zero value for the sum of the fitness measure across the industry.Footnote ¹⁵

To focus directly on the cost impact of the process of differential firm growth, unit cost at each firm is assumed to remain constant during this process and each firm is assumed to operate at full capacity. Industry average unit cost falls as the market share of the lowest cost firms grows. Taking the derivative of industry average unit cost with respect to time yields,

$\left(d{\overline{\text{c}}}_{i,t}/dt\right)={\sum }_{k=1}^n[s_{k,i,t}*(\partial c_{k,i,t}/\partial t)+c_{k,i,t}*(\partial s_{k,i,t}/\partial t)]$(5.9)

With no change in unit cost at any firm, the first term in brackets on the right-hand-side of (5.9) is zero. Imposing this condition and substituting for the change in market share from (5.8) yields,

$\left(d{\overline{\text{c}}}_{i,t}/dt\right)={\sum }_{k=1}^n{s}_{k,i,t}*[c_{k,i,t}*\left(g_{k,i,t}-g_{i,t}\right)]<0$(5.10)

The right-hand-side of (5.10) gives the covariance of unit cost and growth rate across firms, which is negative when differential firm growth is driven by cost-reducing innovation. Metcalfe (Reference Metcalfe1998) demonstrates that the covariance of unit cost and growth rate across firms declines with differential firm growth as market share concentrates in the low-cost firms.

The growth rate of industry supply is given by the weighted sum of growth rates over all firms in the industry, assuming all firms continue to operate and there is no entry. Substituting from the expression for firm growth rates from (5.7), and assuming for simplicity that all firms have the same internal accumulation rate, ${\theta }_{i,t}$ , and the same capital-to-sales ratio, ${\Phi }_{i,t}$ , yields,

$g_{i,t}^S={\sum }_{k=1}^n{s}_{k,i,t}*g_{k,i,t}={\sum }_{k=1}^n\left[s_{k,i,t}\text{*}\left({\theta }_{i,t}/{\Phi }_{i,t}\right)({\text{π}}_{k,i,t-1}-1\right)]=\left({\theta }_{i,t}/{\Phi }_{i,t}\right)({\overline{\text{π}}}_{i,t-1}-1)$(5.11)

Thus, the growth rate for industry supply is proportional to the margin by which the share-weighted average price-cost ratio for the industry, ${\overline{\text{π}}}_{i,t}$ , exceeds one.

Balanced growth in supply and demand is achieved when the growth of supply equals the growth of demand. Decomposing demand growth into underlying growth at a constant price, given by ${\alpha }_{i,t}$ , and the growth of demand due to price change yields,

$g_{i,t}^D={\alpha }_{i,t}+{\eta }_{i,t}(d{p}_{i,t}/dt)\left(1/p_{i,t}\right),{\eta }_{i,t}<0$(5.12)

where ${\eta }_{i,t}$ is the price elasticity of industry demand. Equating (5.11) and (5.12) yields the price-cost ratio required for balanced growth of industry-level supply and demand,

${\overline{\text{π}}}_{i,t-1}=1+\left[{\alpha }_{i,t}+{\eta }_{i,t}\left(d{p}_{i,t}/dt\right)\left(1/p_{i,t}\right)\right]\left({\Phi }_{i,t}/{\theta }_{i,t}\right)$(5.13)

Balancing growth of supply and supply imposes perfect market order on an industry experiencing structural transformation through differential firm growth, which is implausible. The result is a thought experiment demonstrating implications for the industry of firm behavioural routines regarding investments in expanding productive capacity. Each firm implicitly assumes realised profitability is indicative of the profitability of this investment. As is shown next, they are correct as concerns their profitability relative to other firms in the industry, but the profitability of all firms falls over time. Past profitability is an imperfect guide to future profitability.Footnote ¹⁶ Analysis of the implications of differential firm growth without imposing perfect market order is flagged in Chapter 8 as a topic for further research.

Declining average unit cost in (5.10) increases the weighted average price-cost ratio for the industry unless offset by a decline in industry price. Initially, a higher price-cost ratio is required in (5.13) to finance the higher growth in demand with a declining industry price. However, the variance of unit cost across firms declines over time and approaches zero as the market share of low-cost firms approaches one. The rate of cost change in (5.10) converges on zero along with the rate of price change, implying a decline in the rate of growth in demand in (5.12) and in the price-cost ratio required for balancing in (5.13). The fall in the price-cost ratio means price falls faster than the rate of decline in weighted average unit cost.

For each firm, falling price implies a falling price-cost ratio as unit cost is assumed constant at the firm level. The growth rate for each firm declines according to (5.7). Balanced growth of industry supply and demand growth is maintained through the rising share of low-cost firms with relatively high price-cost ratios and growth rates.

Declining industry price with constant unit cost for individual firms means the price-cost ratios for high-cost firms eventually drop below 1, which implies their net investment and growth is negative in (5.7). In the short period, depreciation sets a limit on how much negative net investment can occur in a period. The relationship between the price-cost ratio and firm growth in (5.7) no longer applies, but the course of industry dynamics slows only imperceptibly as the high-cost firms are already relatively small. If the price drops below unit operating cost for these firms, they cease production altogether, effectively exiting the industry, which speeds up market-share dynamics. Obsolescence is the cost of Schumpeterian competition for these firms, their fixed capital, and the specialised skills of their workers. Exit of firms is discussed in more detail in Chapter 6.

Assuming a smooth rate of growth in supply and demand is convenient for analytical purposes but not consistent with capitalist reality. As Schumpeter (Reference Schumpeter1939) argues, innovation-driven growth under capitalism is inherently uneven. Growth of industry demand at constant price is decomposable into a trend component and an annual disturbance,

${\alpha }_{i,t}={\alpha }_i+{\epsilon }_{i,t},$(5.14)

where ${\alpha }_i$ is a trend rate of growth and ${\epsilon }_{i,t}$ is a disturbance term with zero mean and positive variance. How does this affect the dynamics of the market-determined price?

Fluctuations in the value of ${\epsilon }_{i,t}$ in (5.14) create an imbalance between the growth in supply determined by the prior year’s price-cost ratio and the current growth in demand. Price adjusts to balance the predetermined supply growth with the current demand growth. The current weighted average price-cost ratio then moves in the direction of price. If growth in supply follows the path indicated by (5.11), there is lagged adjustment to the temporary change in demand growth.

Lagged adjustment between demand and supply has long been studied in mainstream economics under the rubric of the cobweb model (Kaldor Reference Kaldor1934, Ezekiel Reference Ezekiel1938). Oscillations in price are a feature of this analysis, with the oscillations being damped (explosive) when supply is less (more) responsive to price than is demand. Applying these conditions to equations (5.11) and (5.12) means damped oscillations occur if ${\overline{\text{π}}}_{i,t-1}\left({\theta }_{i,t}/{\Phi }_{i,t}\right)<-{\eta }_{i,t}$ , while explosive oscillations occur if the inequality sign is reversed.

5.2.2 Administered Prices

The analysis of differential firm growth in Metcalfe (Reference Metcalfe1998) focuses on industries with differentiated products and imperfect competition. Prices are determined administratively by a normal pricing routine, where price is set to generate sufficient profit for financing expansion of capacity, taking account of the impact that price has on the growth rate of demand for the firm’s product. The firm offers its full capacity output at that price.

Metcalfe has the growth rate for the firm’s production capacity determined by the same relationship as in (5.7) but expressed with different notation. The k^th firm’s capacity growth rate equals its normal unit profit margin, given by the price-cost ratio minus one, multiplied by what Metcalfe calls the propensity to accumulate, $f_{k,i,t}$ ,

$g_{k,i,t}^S=f_{k,i,t}*({\text{π}}_{k,i,t}-1),f_{k,i,t}>0,$(5.15)

Comparing (5.15) to (5.7), $f_{k,i,t}={\theta }_{k,i,t}/{\Phi }_{k,i,t}$ , so the firm’s propensity to accumulate equals its ratio of net investment to net profit divided by its capital-to-output ratio.Footnote ¹⁷

Following an idea presented in Phelps and Winter (Reference Phelps, Winter and Phelps1970), Metcalfe presents a customer selection model in which the growth in demand for a firm’s product is a linear function of the growth in industry demand and the difference between the firm’s price and the weighted average industry price,

$g_{k,i,t}^D={\alpha }_{i,t}+{\beta }_{i,t}\left[{\overline{p}}_{i,t}-p_{k,i,t}\right],\text{where}{\beta }_{i,t}>0\text{and}{\overline{p}}_{i,t}={\sum }_{k=1}^n\left(s_{k,i,t}*p_{k,i,t}\right)$(5.16)

The firm’s normal pricing routine is to set an administered price that equates the growth in demand in (5.16) to the expansion of production capacity in (5.15), which is,

$p_{k,i,t}=c_{k,i,t}*({\alpha }_{i,t}+{\beta }_{i,t}{\overline{p}}_{i,t}+f_{k,i,t})/\left(f_{k,i,t}+{\beta }_{i,t}\right)$(5.17)

Thus, the firm’s normal price increases with the growth rate for industry demand, ${\alpha }_{i,t}$ , the weighted average price across the industry, ${\overline{p}}_{i,t}$ , and the firm’s unit cost, $c_{k,i,t}$ .

Administered prices determined by (5.17) vary across firms along with variations in unit cost and propensity to accumulate. Simplifying by assuming all firms have the same propensity to accumulate, $f_{i,t}$ , and aggregating across all the firms in an industry,

${\overline{p}}_{i,t}=\left(1+{\alpha }_{i;t}/f_{i;t}\right)*{\overline{c}}_{i,t}$(5.18)

Taking total derivatives of both sides of (5.18), and assuming no change in ${\alpha }_{i,t}$ or $f_{i,t}$ , implies the rate of change in the share-weighted average price equals the corresponding change in unit cost,

$\left(d{\overline{\text{p}}}_{i,t}/dt\right)\left(1/{\overline{p}}_{i,t}\right)=\left(d{\overline{\text{c}}}_{i,t}/dt\right)(1/{\overline{c}}_{i,t})$(5.19)

Industry demand growth in (5.16) doesn’t depend on the rate of change in product price at the industry level simplifies the analysis of dynamics of prices, costs, and market shares through differential firm growth.Footnote ¹⁸ It also suggests an analogy with the replicator dynamics of evolutionary biology. Metcalfe (Reference Metcalfe1998, p. 61) refers to Fisher’s Principle, the Fundamental Theorem of Natural Selection of R.A. Fisher (1930), which ‘states that selection improves average fitness in the population, and that the rate of improvement is equal to the variance of fitness’.

The measure of fitness in Metcalfe’s application of replicator dynamics is the relative growth of the firm. Application of Fisher’s Principle implies that the simple average growth rate of firms across an industry falls as the variance of firm growth rates declines. However, the growth rate for industry supply remains equal to the growth of industry demand, with the rising market share of low-cost firms maintaining a constant weighted average growth rate for industry productive capacity.

Of more direct relevance to analysing price dynamics is the Secondary Theorem of Natural Selection, which Metcalfe (Reference Metcalfe1998, p. 63) states, ‘Consider any behavioural trait, then the rate of change population average of this trait and fitness across the population equals the covariance between that trait and fitness across the population’. Equation (5.10) provides an illustration of this theorem, with the rate of change of industry average unit cost equal to the covariance between firm unit cost and firm fitness (measured by the difference between its growth rate and the industry average).

Equation (5.10) is a generic expression that applies to any analysis of differential firm growth. Substituting from (5.16) through (5.18) into (5.15) uses the growth and pricing assumptions of the administered pricing analysis to derive an expression for firm growth in terms of the difference between its unit cost and the industry weighted average as follows,

$g_{k,i,t}^S={\alpha }_{i,t}+[{\overline{c}}_{i,t}-c_{k,i,t}\left]\left[({\beta }_{i,t}*f_{i,t}\right)/\left(f_{i,t}+{\beta }_{i,t}\right)\right]$(5.20)

Further substituting from (5.20) into the generic expression in (5.10) for the rate of change in industry average unit cost yields,

$\left(d{\overline{\text{c}}}_{i,t}/dt\right)={{\sum }_{k=1}^n[s}_{k,i,t}*c_{k,i,t}*\left(c_{k,i,t}-c_{i,t}\right)\left]\left[({\beta }_{i,t}*f_{i,t}\right)/\left(f_{i,t}+{\beta }_{i,t}\right)\right]<0$(5.21)

The first term in brackets in (5.21) is the weighted variance of unit cost across the industry. Thus, the variance of unit cost across firms is driving the differential firm growth process, which results in rising average efficiency as well as declining weighted average unit cost and price.

Metcalfe (Reference Metcalfe1998, p. 51) shows the share-weighted variance of price across firms in the industry is proportional to the share-weighted variance in unit cost across the firms,

$Va{r}_s\left(p_{k,i,t}\right)={\sum }_{k=1}^n[s_{k,i,t}*{\left(p_{k,i,t}-{\overline{p}}_{i,t}\right)}^2]=Va{r}_s\left(c_{k,i,t}\right)*{\left[f_{i,t}/\left(f_{i,t}+{\beta }_{i,t}\right)\right]}^2$(5.22)

As the selection mechanism of differential firm growth reduces the variance of unit cost, the share-weighted variance of prices falls in tandem. Thus, the process of differential firm growth is degenerative with administered prices, as is the case for market-determined prices in Section 5.2.1. Heterogeneity is devoured by the process that drives differential growth.

5.3 Post-Keynesian Parallels

Overlap between the administered pricing analysis of the previous section and the work of Kalecki (Reference Kalecki1971) is readily apparent in comparing the normal pricing routine in (5.17) to the Kaleckian pricing routine in (3.3). A key difference is the growth of industry demand is included in price determination in (5.17), whereas Kalecki treats the pricing coefficients m and n as determined by a range of factors, including industry concentration, the degree of sales promotion, overhead costs as a proportion of total costs, and the strength of trade unions.

Kaleckian and other post-Keynesian pricing models have firms set administered prices with the expectation of sales usually being below production capacity, whereas in pricing analysis in the previous section assumes full capacity utilisation. Firms operate with excess capacity to enable them to meet fluctuations in demand and to dissuade entrants. Fluctuations in demand for the industry’s product therefore don’t affect administered prices, avoiding the complication of oscillating investment and prices encountered when prices are market-determined.

Despite having a normal pricing routine based on full capacity operation, Metcalfe (Reference Metcalfe1998, p. 93) concurs with the post-Keynesian treatment of fluctuating demand, stating, ‘in the first instance, the impact of turbulent conditions will be reflected in departures from full capacity operation, in excess order books or unfilled capacity.’ Demand fluctuations impact realised profits but there are no price oscillations in the analysis for administered prices in Section 5.2.2. Only expected demand growth impacts the growth of productive capacity determined by (5.16) and the administered price determined by the normal pricing routine in (5.17).

Linking of price determination to the internal financing of investment in Section 5.2.2 has parallels in post-Keynesian analysis by Eichner (Reference Eichner1973, Reference Eichner1976 and Reference Eichner1985), Wood (Reference Wood1975), and Harcourt and Keynon (Reference Harcourt and Kenyon1976). In each case the firm administratively determines a price, taking account of both the positive impact of a higher price in providing finance for the expansion of production capacity and the negative impact of a higher price on consumer demand for its product. The distinctive features of each of these works and their links to the broader price theory literature are discussed in Melmiès (Reference Melmiès2022) and Bloch and Kriesler (Reference Bloch and Kriesler2023).

Post-Keynesian pricing analysis generally doesn’t deal directly with the phenomenon of differential firm growth. An exception is the analysis of Steindl (Reference Steindl1976). Steindl analyses the pattern of competition driven by internal accumulation for an industry with heterogeneous firms. As in the evolutionary analyses of Nelson and Winter (Reference Nelson and Winter1982) and Metcalfe (Reference Metcalfe1998), Steindl has low-cost firms growing faster than high-cost rivals due their greater profits and investment through internal accumulation. With a persistent cost advantage, the low-cost firms grow faster than industry average leading to rising market shares, such as for the entrepreneurial firm in (5.1).

Steindl suggests the low-cost firms engage in aggressive price and selling competition to continue growing faster than the industry, which eventually drives many high-cost firms from the industry. When the share of high-cost firms becomes negligible, the low-cost firms recognise their collective interest in reducing aggressive competition and the growth of their productive capacity. After this, low-cost firms use high profits to pay off debt or increase returns to shareholders and management.Footnote ²⁰

Steindl’s analysis is distinguished from the evolutionary analysis of differential firm growth in Section 5.2.2 by incorporating changes in investment and pricing routines of low-cost firms in response to rising industry concentration. Changes in pricing and investment behaviour are changes in routines, which are compatible with evolutionary analysis. Markey-Towler (Reference Markey-Towler2016) explores the constraint market demand imposes on firm growth and how the choice of pricing routine is related to this constraint. Almudi et al. (Reference Almudi, Fatas-Villafranca, Palacio and Sanchez-Choliz2020) consider the impact that consumer learning has on demand and how firms react by changing their pricing routines. Integrating changes in firm pricing routines into analysis of the evolution of industry structure is a useful direction for future research in evolutionary price theory.

5.4 Summary

An economy experiencing innovation-driven growth is an economy where prices are in motion. Current prices coordinate buyers and sellers, creating market order, while providing only imperfect guides to future prices on which the profitability of investments depend. Heterogeneity across producers is common, even for similar or identical products. Movements in prices over time reflect the outcome of the contest between rival producers with their different products, processes, and organisations, a contest heavily influenced by the institutional environment in which firms operate.

Under capitalism profits are a market test for firms. They also provide firms with access to finance for expansion. High profits mean firms with superior products, productivity, or organisation can demonstrate their fitness by expanding relative their inferior rivals. The two models of differential firm growth developed in this chapter show how cost heterogeneity translates into differential firm growth, price dynamics, and the evolution of industry structure. One model assumes individual firms are price takers, with price determined by the interaction of supply and demand in the market. The other model assumes firms set prices through administrative routines designed to balance the growth of their production capacity with the growth of demand for their product.

Replicator dynamics relating the change in weighted average unit cost to the share-weighted covariance of firm cost differences and growth differences, as shown in (5.10), apply with both market-determined and administered prices. The covariance is negative with lower-cost firms growing faster, so weighted average unit cost declines over time. Balance between the growth of industry supply and demand means prices decline as weighted average unit cost declines, whether through administrative design or through coordination by market price.

6.1 Boundaries of Firms and Industries

Analysis in Chapter 5 associates a firm with a particular market in competition with a group of firms that form an industry. Further, it is assumed each firm operates a single production process that expands or contracts at constant unit cost. All this is convenient for an analysis focusing on the market as a selection environment for competition among the heterogeneous group of firms in the industry, with creative destruction as the driver of industry evolution and price dynamics.

Diversifying into new markets or developing new products allows firms to continue internal accumulation without engaging in aggressive competition to increase their share of existing markets. Restless knowledge ensures a steady supply of new ideas, which may be captured as innovations by firms able to search, seize, and secure the opportunities. Teece (Reference Teece, Dopfer, Nelson, Potts and Pyka2024) argues this requires dynamic capabilities for going beyond established routines.

If firms are not confined to specific products or markets, what meaning remains for the concept of an industry? An answer starts with Marshall’s (Reference Marshall1920) analogy between firms in an industry and trees in a forest. There are both complimentary and competitive relationships among firms engaged in related activities. They share overlapping sets of customers and suppliers, as well as connections through industry associations, training colleges, specialised consultancies, and the like. Connections across a group of firms stronger than connections to firms outside the group define an industry.

Mainstream analysis has emaciated Marshall’s analysis of the firm and its relationship to the industry. Pigou (Reference Pigou1920) replaces Marshall’s heterogeneous firms (trees in the forest) with the equilibrium firm, while Robinson (Reference Robinson1931) insists the equilibrium firm is an optimising firm. Heterogeneity is ruled out.

Individuals and firms are at the core of evolutionary development, but not as the isolated economic agents assumed in methodological individualism. Non-market interactions are central to the process that leads to development by connecting the specialised knowledge of individuals through the exchange of information. As discussed in Chapters 2 and 3, these interactions occur in multiple contexts, including among individuals as consumers, between individuals within firms, among firms within industries. Firms and industries also have interactions with governments, universities, and other social organisations. Discussion in this section explores some of these interactions and their implications for understanding the changing boundaries of firms and industries.

6.2 Entry and Exit

Structural change in terms of changes in the relative size of industries is a standard feature of evolving economies along with the creation of new industries and the disappearance of old industries. Sharply contrasting analyses of the relationship between firm size and industry size are provided by evolutionary and neoclassical economics. In the analysis of differential firm growth in Chapter 5, changes in industry size are accommodated through adjustment of existing firm sizes (fit firms growing and unfit firms stagnating or shrinking). No entry or exit of firms occurs. In neoclassical economics with long-run perfectly competitive equilibrium, changes in industry size occur through a change in the number of firms, each firm operating at the same output level with price equal to minimum long-run average cost (ignoring problems associated with fractions of firms). No change occurs to the size of existing firms.

Entry and exit of firms are important phenomena and deserve to be analysed constructively in the theory of price determination, whether evolutionary or neoclassical. Sporadic attempts at a theory of entry by neoclassical economists over the years include Marshall’s (Reference Marshall1920) analysis of long-run supply based on the representative firm, models of barriers to entry by Bain (Reference Bain1956) and Sylos-Labini (Reference Sylos-Labini1962), and game-theoretic models of firm strategy toward technology and marketing by Sutton (Reference Sutton1991 and Reference Sutton1998). These attempts have been met by subversion (witness the fate of Marshall’s representative firm), counterattack (for example, Baumol et al. (Reference Baumol, Panzar and Willig1982) attacking models of barriers to entry with the concept of contestable markets), and, mostly, indifference. Determining the amount of entry and exit into an industry by equating price to minimum average cost is an article of faith for the mainstream, rather than the result of a theory of entry.

Entry and exit of firms can be incorporated into the analysis in Chapter 5. Exit of firms is a straightforward extension. When industry price falls below a firm’s unit cost, the firm’s operating surplus (the gap between revenue and operating cost) is below depreciation charges and net investment is negative. If the firm’s operating surplus is below its fixed expense (interest on loans, lease payments, and the like), it becomes insolvent. If price drops below operating cost per unit of output, production ceases even without fixed expenses.

Entry of firms is an indirect effect of differential firm growth when entry into new industries is an outcome of diversification efforts as discussed in Section 6.1.3. The presumption in Chapter 5 is that the knowledge and organisational capability required for operating in the industry is limited to existing firms. However, these obstacles to operating in the industry aren’t insurmountable. Firms in the supply chain may diversify upstream or downstream using knowledge and capabilities from their network connections. Employees of existing firms are another source of new entrants, especially employees coming from low-cost firms.

An entrant’s survival is far from assured. An immediate threat to its survival occurs if actual unit cost exceeds the expected level or actual price received is below expected price, leading to lower profits than anticipated. Afterward, new firms are exposed to the same competitive selection environment as existing firms. Unless the entrant’s unit cost is near or below that of the lowest cost firms, falling prices due to high growth of the lowest cost firms eventually result in prices below the entrant’s unit cost. Creative destruction is an existential threat to entrants and established firms alike.

Once an entrant commits its financial and organisational resources to an industry it is unlikely to be able to exit without experiencing a substantial loss on the value of these resources. The inability to exit without loss constitutes a barrier to entry for an industry experiencing creative destruction. Potential entrants into such an industry only enter if current average product price for the industry is substantially above the entrant’s expected unit cost (Eaton and Lipsey Reference Eaton and Lipsey1980).

Metcalfe (Reference Metcalfe1998) adds entry to the analysis of differential firm growth by assuming a potential supply of new firms. A firm enters only when industry price is at a minimum threshold above its expected unit cost. Expected unit cost is heterogeneous across the potential entrants, for reasons similar to those causing unit cost to differ across existing firms. Entry occurs at a finite rate that increases with the industry average price, as more firms find the price exceeds their threshold.

Entry modifies the dynamics associated with differential firm growth by existing firms. Entry adds to production and puts downward pressure on price. Profits of the existing firms are lower, which slows the growth of their production capacity. As Metcalfe (Reference Metcalfe1998) argues, what matters to industry price and the dynamics of market shares of established firms is the share of industry sales taken by entrants as a group, which is generally small.

These observations about the drivers of entry and exit are far from providing a satisfactory theory. In the context of established industries, they are ancillary to the theory of differential firm growth in determining price dynamics. However, the entry process is fundamental for understanding the evolution of new industries, and this involves institutional and sociological influences as well as economic (see Gustafsson et al. Reference Gustafsson, Jääskeläinen, Mauls and Uotila2016 for a discussion). Bloch (Reference Bloch2018c) discusses both the mainstream and evolutionary economics literature on innovation and the evolution of industries.

6.3 Inter-industry Linkages

Industries are linked by the flow of products, as well as the flow of firms, across industry boundaries. When the output of one industry is used as an input by another industry, the revenues of the first industry are costs for the second. Price movements in input-supplying industries thus flow through to costs and prices throughout the economy. These linkages in price are discussed in Section 6.3.2. First, a mechanism is required for summarising the movement of unit cost and price within an industry when firms are heterogeneous, with the representative firm proposed as the mechanism.

6.3.2 Inter-industry Linkages in Prices

Industries are linked through the flow of intermediate products, with services, construction, and durable equipment being outputs of one industry and inputs to another. Input-output analysis developed by Leontief (Reference Leontief and Leontief1986) provides an empirical framework for linking production flows across industries. Neoclassical economics builds on this framework using Walrasian general equilibrium theory. However, as discussed in Chapter 4, equilibrium theory is inappropriate for use in evolutionary analysis as the requirements of optimising behaviour along with externally determined technology, preferences and resources are incompatible with analysing economic evolution.

Sraffa (Reference Sraffa1960) examines the requirements for reproduction in an economy where commodities are produced through combining commodities and labour. In the tradition of long-period analysis of classical economics, the price for each commodity is equal to its unit production cost calculated using the corresponding prices of all commodities used as inputs along with a uniform wage rate and rate of profits across commodities. Roncaglia (Reference Roncaglia1978, p. 16, apostrophes in original) points to the omission of any requirement for equilibrium as a feature distinguishing Sraffa’s analysis from neoclassical analysis, ‘The fact that Sraffa never talks of an “economic equilibrium” or of “equilibrium prices” in relation to his system should also be emphasized.’

Sraffa shows for any given structure of technical requirements of commodity inputs and labour for producing each commodity there is a corresponding unique set of reproduction prices for any given wage rate or profit rate. There is also a monotonic inverse relationship between the wage rate (assuming homogenous labour) and the rate of profit. Thus, if one rate is set the value or the other is uniquely determined.

Roncaglia (Reference Roncaglia2006) distinguishes three orientations by Sraffa scholars towards classical economics. First is the Ricardian reconstruction of Pasinetti (Reference Pasinetti1981), second is the Marxian reconstruction of Garegnani (Reference Garegnani, Bharadwaj and Schefold1990), and third is the Smithian reconstruction of Sylos-Labini (Reference Sylos-Labini1984). Roncaglia (Reference Roncaglia2006, p. 466) notes the Smithian approach has an advantage in terms of flexibility regarding prices and profits, ‘while the notion of surplus retains a central role in economic analysis, the functional relationships connecting production prices and income distribution lose their role as the central pillar of economic theorising’.

Bloch (Reference Bloch2018b) provides an evolutionary analysis of inter-industry linkages in prices using a modification of Sraffa’s framework following a Smithian approach. The evolutionary analysis uses the following system of equations in matrix notation,

$p=\pi u=\left({\left(I–\pi A\right)}^{-1}\right)\pi \gamma w,$(6.1)

where p is the vector of commodity prices (the i^th element, p_i, is the price of the i^th commodity), $\pi$ is a diagonal matrix of price-cost ratios (only diagonal elements are non-zero and the i^th diagonal element, π_ii, is the price-cost ratio for the i^th industry, u is the vector of unit production cost for all commodities, A is the matrix of input-output coefficients,γ is the vector of labour requirements and $w$ is the vector of share-weighted average wage rates, with elements of w_i. Elements of the matrix A give the share-weighted direct requirements for producing each commodity, where a_ij gives the direct requirements of commodity j used in producing one unit of commodity i. Elements of the vector, γ, give the share-weighted average direct labour requirement for producing each commodity, with elements of γ_i.

In adapting Sraffa’s framework for evolutionary price theory, the vectors and matrices used in (6.1) are values for the redefined version of Marshall’s representative firm rather than assuming a single production process applying to all producers as in Leontief (Reference Leontief and Leontief1986) or Sraffa (Reference Sraffa1960). In Section 6.1.5 the redefined representative firm has share-weighted average values of price, unit cost and the price-cost ratio. Implementing this interpretation in the system of equations in (6.1) implies average unit cost is determined by share-weighted average technical coefficients and a share-weighted average wage rate for the industry, while the share-weighted average price equals this share-weighted average unit cost multiplied by the share-weighted average price-cost ratio.

Using the redefined representative firm with share-weighted average industry values of technical coefficients, price-cost ratio, and wage rate has the advantage that flows of revenues and cost add up across the economy. The system of equations in (6.1) is interpretable as a system of accounting relationships in which the price in each industry is equated to the average unit cost of production times an average price-cost ratio. This interpretation fits with the Smithian reconstruction of Sraffa as noted by Roncaglia (Reference Roncaglia1978), although shifting the focus of analysis from long-period analysis of production prices and income distribution to the analysis of the trajectory of costs, profitability, and prices occurring in the process of creative destruction.Footnote ²⁴

Differential firm growth with creative destruction leads to changes in average technical coefficients, unit costs, price-cost ratios, and prices. These changes occur because of changing the market shares of producers, even without unit cost changing at individual producers. Diffusion and adaptation to major innovations occur over long periods of time, witness the decades taken for diffusion of railways, mass production of motor vehicles, or automation in food manufacturing.

As, and the word, price changes within an industry, the system of equations in (6.1) provides a framework for tracing impact throughout the economy.Footnote ²⁵ Industries using inputs from an industry experiencing creative destruction are most directly impacted. If the product of the industry experiencing creative destruction is a basic commodity as defined by Sraffa, prices in all industries are impacted without any change in their wage rate or price-cost ratio. Otherwise, the range of affected industries is limited but extends beyond the initial industry, aside from the case of commodities that aren’t used in the production of any other commodity. Quantities are affected as well as prices, with industries whose relative prices fall expanding and those with higher relative prices contracting as suggested by the analysis of orderly markets in Chapter 4.Footnote ²⁶

6.4 Prices in the Long Run

A theory for determining prices in the long run has been the holy grail of economics.Footnote ²⁷ Classical economic theory of natural prices relates the long-run relative price of a commodity to its cost of production. Ricardo ([Reference Ricardo1821] Reference Ricardo1973) argues this means long-run relative prices approximate their relative labour content, including both direct and indirect labour. Marx ([Reference Marx1887] Reference Marx1954) shifts focus slightly and argues long-run relative values depend solely on relative direct and indirect labour, but then has trouble converting long-run relative labour values into long-run relative prices.Footnote ²⁸

Sraffa (Reference Sraffa1960) uses the framework of production of commodities by means of commodities to clarify confusions of classical theory, including demonstrating an invariable standard of value in the form of the standard commodity. He also demonstrates the invariable standard of value is unique to each set of technical coefficients, so technical change, unless it is balanced across all commodities, changes the standard commodity. Thus, comparisons over time for economies experiencing structural change inevitably must be based on variable standards of value.

Neoclassical economists object to purely cost-based determination of prices in the long run, especially the notion that prices are determined by labour values. They argue unit cost generally depends on the scale of production, so demand as well as supply conditions affect relative prices in the long run. Also, they introduce marginal analysis for determining equilibrium prices. This has led to a theory of long-run relative prices based on intertemporal perfectly competitive general equilibrium with rational expectations. These prices are stable with respect to random disturbances. They only change with unexpected external influences. Very neat, but of questionable relevance for analysing an evolving economy, where prices are determined by endogenous processes of innovation and structural change.

Marshall (Reference Marshall1920) introduces the representative firm in analysing long-run supply conditions for an industry, which he then combines with long-run demand conditions to determine long-run relative prices using his famous scissors analogy. The unit cost for the representative firm includes the cost of use of durable plant and equipment along with a normal return on invested capital. Marhsall argues entrants look to the representative firm as indicative of what they will experience after entry, with firms entering the industry whenever price exceeds the unit cost of the representative firm. Firms exit whenever their unit cost is continuously below price. Marshall concludes the long-run supply price for an industry equals the unit cost for the representative firm.

Evolutionary price theory doesn’t offer a theory of price in the long run as an equivalent to the natural price of classical theory, Marshall’s long-run supply price, or the intertemporal general equilibrium price of modern neoclassical theory. Instead, it provides a theory of price determination in the process of the introduction and diffusion of innovations through creative destruction, which leads to a trajectory of prices over time. There are several important consequences.

First, rather than a fixed point or central tendency that market prices fluctuate around as in classical and neoclassical economics, there is a lower boundary to the trajectory of prices. Firm heterogeneity leads to differential firm growth and industry average price declines over time. In Chapter 5, the aspect of heterogeneity analysed is variation in unit cost, with the rate of price decrease depending on the variance of unit cost across firms. Without further innovation-induced changes in unit cost at the firm level, differential growth reduces variance of unit cost and the rate of price change declines over time as price approaches its lower boundary.

Second, evolutionary price theory occurs in historical time, not the logical time of classical or modern neoclassical price theory. The process of creative destruction is irreversible. It is also path dependent with stochastic events, such as innovation or entry of new firms, interrupting the decline in firm heterogeneity. Because innovation and entry are successful only to the extent that the entrant or innovator is fitter than average for the industry, they tend to enhance rather than reverse the downward trajectory of costs and prices.

Third, differential firm growth is driven by profit differentials across firms, with internal accumulation of the fittest firms driven by high relative profitability. The natural price in classical economics equals the unit cost of reproduction based on uniform profit and wage rates across firms. A uniform normal rate of return is included in both Marshall’s long-run supply price that equals the unit cost of the representative firm and in neoclassical long-run perfectly competitive equilibrium price that equals minimum unit cost using best-practice technology. Profit in evolutionary price theory is driven by creative destruction, a degenerative process, rather than representing a payment for the use of capital as in classical and neoclassical price theory.

Fourth, the process of creative destruction takes time, more time than implied by Marshall (Reference Marshall1920) in his distinction between short-run and long-run adjustment processes. Marshall associates the difference between the short and long runs with the time required to expand or contract production capacity to meet demand. The analysis is Chapter 5 presumes firms build capacity ahead of demand over the interval between periods. Yet, creative destruction continues indefinitely, limited by the speed with which the fittest firms can expand their market shares at the expense of less fit rivals. As the heterogeneity across firms declines, the speed of adjustment declines so heterogeneity disappears asymptotically at the infinite horizon. Innovation and entry add to variety and, thereby, interrupt the process of adjustment.

6.5 Summary

Part III focuses on aspects of evolutionary price theory that extend beyond an isolated consumer, firm, or industry. The emphasis in this chapter is on influences that link industries together, starting with restless knowledge. Restless knowledge generates opportunities for innovation that extend beyond traditional boundaries of the firm and industry. Firms with dynamic capabilities exploit these opportunities, making industry boundaries porous in an evolving economy.

Entry and exit of firms affect the dynamics of market shares and prices in industries undergoing creative destruction, which impacts results from the analysis of differential firm growth in Chapter 5. Except for new industries, the impact is unlikely to be substantial due to the small share of industry output accounted for by entrants or exits. A theory of entry and exit suitable for evolutionary analysis is a work in progress.

Price changes spread through the economy through the flow of products between industries. For purposes of tracing these impacts, the industry is represented by a firm with share-weighted average values of technical production coefficients, wage rate, unit cost, price-cost ratio, and price. A system of equations with these representative firm values for each industry provides the mechanism for calculating the direct and indirect impacts of changes in the price for the representative firm in one industry on costs and prices for representative firms in all other industries.

Evolutionary price theory is a theory of the movement of prices through time rather than a theory of the structure of prices in the long run as in classical and neoclassical economics. Prices move with the process of differential firm growth. The process occurs slowly over time, is irreversible and path dependent. The process is degenerative, with shrinking market shares of unfit firms consuming the heterogeneity that drives differential firm growth. However, entrepreneurial action based on restless knowledge, creativity, and imagination ensure variety is replenished and development from within continues indefinitely.

7.1 Entrepreneurs, Credit, and Inflation

Entrepreneurs are essential to Schumpeter’s theory of innovation-driven growth. They are distinguished by their imagination and by their sociological character of leadership, seeing opportunities invisible to others and convincing others of the viability of their vision. Bankers are also essential as entrepreneurs don’t own means of production and require cooperation from bankers to finance their innovations.

Schumpeter ([Reference Schumpeter1934] Reference Schumpeter1961, Chapter 3) notes his theory of economic development, with its emphasis on structural change through entrepreneurship, leads him to a heretical position on the role of credit and money. In particular, he rejects the view that money is a veil with no real influence on the course of economic development. Instead, the creation of credit that expands the money supply is necessary for entrepreneurs to acquire the means of production, diverting these inputs from established firms for use in innovative production activities.Footnote ²⁹

Schumpeter’s heresy on credit and money extends to his treatment of capital. Schumpeter (Reference Schumpeter1939, p. 129) treats capital ‘as an accounting concept – as measuring in terms of money the resources entrusted to a firm’. Bankers serve as social accountants and gatekeepers through their granting of credit to entrepreneurs. Projects that are deemed able to return principal and interest are funded and become capital.

How can it be that financing innovation generates capital even though there are no additional means of production, only a shift of means of production from established producers to innovators? Innovation is an out-of-equilibrium process. The value of the innovation derives from the surplus between the stream of future revenues from sale of the innovative product and the stream of costs, including expenditures on the initial means of production and subsequent expenditures. For a successful innovation, this surplus exceeds the surplus that would have been obtained from the use of the means of production by established producers. Thus, the value of the capital generated by financing innovation is validated ex post by the ability of entrepreneurs to repay banks with interest and still obtain profits. If the innovation fails to meet this market test, the entrepreneur is liquidated, the bank suffers a loss, and the capital is invalidated.

Schumpeter (Reference Schumpeter1939, p. 123) defines interest in purely monetary terms: ‘Interest is a premium on present over future means of payment.’ He argues neither abstinence nor roundabout production is required to explain interest. Instead, entrepreneurial profits provide a logically sound, albeit transitory, source to pay interest. Interest is in a sense a tax paid by entrepreneurs to banks for obtaining access to means of production in advance of owning a share of society’s resources. With innovations fully absorbed at the end of the business cycle, entrepreneurial profits are zero. Nothing is left with which to pay interest, so interest logically equals zero in a stationary state.Footnote ³⁰

The credit banks extend to entrepreneurs adds to the flow of expenditure in the economy, especially expenditure on the means of production, including labour, raw materials, and durable equipment. Prices of the means of production rise, increasing production cost for new and established producers alike. Wages rise, with these increased costs passed through to increased prices for intermediate and final products as indicated in the analysis of Section 6.3.2. The aggregate price level as measured by an index of weighted elements in the price vector then rises.Footnote ³¹

Inflation associated with an upsurge in innovations is temporary. Schumpeter has entrepreneurs using profits from successful innovations to repay their loans. When entrepreneurs repay their loans, money supply falls as does aggregate nominal income, a process Schumpeter calls autodeflation. Additional output from the entrepreneurs more than replaces the output lost from established producers, so prices fall on average, contributing to falling nominal income.Footnote ³² Thus, Schumpeter concludes that there is a downward trend for the price level over the long period.

Minsky (Reference Minsky, Heertje and Perlman1990), Knell (Reference Knell2015), and Callegari (Reference Callegari2018) argue changes in the financial market institutions mean the simple relationship between credit and innovation posited by Schumpeter’s analysis needs updating to explain cyclical behaviour in a modern economy. Schumpeter ([Reference Schumpeter1950] Reference Schumpeter1976) acknowledges the start of this institutional shift, noting the enhanced role of large corporations in the innovation process. A large corporation shifting means of production from an established line of business to an innovative activity doesn’t require access to external finance, nor does its innovative activity necessarily impact wage rates. Still, expansion of the innovative activity requires displacing established production and disruption of markets, except in the unlikely event that the large corporation is cannibalising its own markets. Thus, the inflationary and deflationary effects of credit expansion and contraction may be avoided, while the downward pressure on the price level over the long period remains.

Further discussion of Schumpeter’s analysis of movements in the price level is postponed to Section 7.3. There, flaws in Schumpeter’s explicit and implicit assumptions about micro-level behaviour and meso-level adjustment processes are identified and an integrated framework for analysing movements in prices across the economy is presented. First, the existence of long waves in the intensity of innovative activity and the links between these waves and the reliability of price information is considered in the following section.

7.2 Waves of Innovation

As discussed in Chapter 5, the intensity of innovation activity is tied to the reliability of price information in Schumpeter’s endogenous process of growth and structural change. When current price information provides a reliable guide to future prices, entrepreneurs can accurately calculate the profitability of potential innovations and obtain credit from banks to finance those innovations. Innovations and their diffusion through creative destruction disrupt markets, making price information less reliable and further innovations are discouraged.

Feedback between the reliability of price information and the intensity of innovation is at the core of Schumpeter’s theory of the business cycle. Reliable price information leads to a wave of innovations, a wave of innovations increases price growth but disturbs the price system. Uncertainty about future prices depresses innovative activity. Low innovative activity stabilises output and prices, reducing uncertainty about future prices. Repetition of the sequence over time creates waves of innovative activity that are the driving force of cycles in price and output growth.

The analysis in Chapter 5 relates the rate of price adjustment to the variance of cost differentials across producers. If the variance of cost differentials were known, insert, an estimate of, the rate of price adjustment consistent with the adjustment process could be calculated. However, this type of information is unlikely to be available to market participants. Also, as Steindl’s (Reference Steindl1976) analysis of price adjustment to cost differentials demonstrates, the price adjustment process may be erratic, with periods of price stability interspersed periods of intense price competition. The price of computer processing power has fallen dramatically over recent decades, but the rate of decline has been far from steady.

Price adjustments in a single industry spread through the economy in the analysis in Chapter 6. Price adjustments directly impact downstream industries through the price of intermediate inputs in the input-output matrix in (6.1). Entry and exit of firms may also impact prices through changes in price-cost margins in affected industries. Structural changes in the shares of various industries in aggregate production occur in tandem with changes in the structure of prices for individual industries.

The unknown dynamics of product prices in industries undergoing diffusion of innovations together with uneven pass through to cost and prices in other industries create uncertainty regarding the difference between cost and revenue for production processes throughout the economy. The sporadic nature of the process of entry and exit adds to uncertainty. Uncertainty of this sort may not have great impact on investments by established firms in their ongoing activities because they have ongoing commitments to employees, leases, loans, and customers. However, entrepreneurs who require access to external finance have difficulty developing business plans that avoid unacceptable uncertainty for their financiers.

Schumpeter’s (Reference Schumpeter1939) Business Cycles expands the business cycle theory beyond the primary cyclical activity. Innovative activity induces secondary expansion in non-innovating industries. This secondary expansion depends on boom conditions to be profitable, which disappear when innovative activity declines. Also, euphoria in the boom leads to banks extending credit to speculators, which means a downturn in innovative activity results in financial crisis as well as liquidation of unsustainable secondary expansion. Secondary phenomena add phases of depression and recovery to Schumpeter’s original two phases of prosperity and recession.

Schumpeter bases the duration of the longest business cycle on a historical analysis of periods of innovation-driven growth in the world economy, settling on a length of fifty to sixty years in line with the work the Kondratieff (Reference Kondratieff1935). He provides beginning and end dates for two long cycles, the Industrial Revolution from 1787 to 1842 and Railroadization from 1843 to 1897, as well as a beginning date, 1898, for a third long cycle of Electrification. He adds shorter cycles of about nine years duration due to the time taken to set up new industrial plant and equipment, Juglar cycles, and inventory cycles of about three years duration, Kitchin cycles. Then he suggests all cycles overlap if the Juglar cycles are 38 months (three and sixth years), the Juglar cycles are 114 months (9 and a half years), and Kondratieff cycles are 57 years in length.

Schumpeter’s (Reference Schumpeter1939) Business Cycles received mixed reviews. Economic historians were generally positive (Innis Reference Innis1940, Rosenburg Reference Rosenberg1940). However, economists tended to be critical, with Kuznets (Reference Kuznets1940), a leading business-cycle analyst of the time, commenting unfavourably on the relationship between the theoretical model and the empirical content. Bloch (Reference Bloch2018a, Chapter 3) provides a detailed discussion of Schumpeter’s business cycle theory and the reactions it evoked.

Despite the critiques and lack of immediate adherents to Schumpeter’s explanation for Kondratieff cycles in innovative activity, his notion that the frequency and impact of innovations follows a wave-like pattern over time has been gaining popularity (Tylecote Reference Tylecote1992, Freeman and Louçã Reference Freeman and Louçã2001, Perez Reference Perez2002, Lipsey et al. Reference Lipsey, Carlaw and Bekar2005, Louçã Reference Louçã2021). While these advocates of waves of innovation generally cite Schumpeter as their inspiration, they tend to emphasise technological, financial, or institutional foundations for the wave-like movement. In doing so, they overlook Schumpeter’s distinction between invention and innovation. They also overlook the endogenous mechanism involved when reliable price information helps overcome the barriers to turning inventions into innovations.

The connection between waves in innovative activity and movements in prices suggested by Schumpeter occurs regardless of the causal mechanism explaining the waves. Whether or not the reliability of price information is the cause of an upsurge in innovation, innovations require the shift of means of production from established producers to innovators and then the shift of expenditure from the products of established firms to those of the innovators. As Hayek (Reference Hayek1935) explains, the price mechanism is key to these structural changes in inputs and outputs across the economy.

7.3 Movements in the Price Level and Price System

Schumpeter (Reference Schumpeter1939) distinguishes between the price level and the price system. The price system refers to the relation of prices of individual products to each other, which is the subject of analysis in Chapter 6. For Schumpeter, the price level is a social construct relating the standard unit of account (money) to the full array of consumer expenditures. Schumpeter (Reference Schumpeter1939, p. 454) warns, ‘It is also easy to see that changes in the level can in practice hardly ever come about except by way of changes in the price system – even as changes in the system in practice hardly ever come about without enforcing a change in the price level.’

7.4 Summary

Motion, rather than equilibrium, is at the core of understanding movements in economic aggregates in a developing economy. Schumpeter identifies waves of innovation as the driving force behind growth and development in capitalist economies. Associated with these waves of innovation are movements in the price system and the price level. The price system is altered through changes in relative prices, the structure of production, and the pattern of consumer expenditure. Aggregate price indexes obscure changes in the price level due to the changing price system, undermining meaningful quantitative comparisons over long periods for aggregate output, the real wage, and the material standard of living.

Mainstream theory is presented as providing a logically consistent framework for economic analysis, including the analysis of economic change. Sraffa (Reference Sraffa1960) demonstrates there is no invariable standard of value for making logically consistent comparisons of prices when production processes are changing, which invalidates comparisons of aggregate output except for the case of steady-state growth. Steady-state growth with an unchanging production structure is implausible for an evolving economy. Mainstream price theory sweeps these contradictions under the carpet.

8.1 Summary

The theory of price determination outlined in this Element is designed to be consistent with presumptions of evolutionary economic analysis. These presumptions include an open-system ontology in contrast to neoclassical economics with its closed-system ontology. Also, the constructed theory is based on co-evolution of micro, meso, and macro elements, which contrasts to the methodological individualism underpinning the microfoundations approach of mainstream economics. Further, the theory incorporates endogenous structural change with historical specificity and path dependence, contrasting to universality and time reversibility in mainstream economics.

Schumpeter’s work on economic development is foundational to the evolutionary theory of price determination. To this are added ideas from evolutionary, behavioural, and post-Keynesian economists on the rules and routines used by consumers and firms. Marshall’s analysis provides a framework for analysing the relationship between firms and industries, especially in orderly markets. Evolutionary and post-Keynesian analyses of differential firm growth provide guidance for the theory of price determination in disrupted markets. Inter-industry linkages of prices are built upon the work of Sraffa on production of commodities by means of commodities.

In Chapter 2, consumers are assumed to behave according to rules or routines, which contrasts to the optimising behaviour assumed in mainstream economics. Following rules or routines is a reasoned reaction to the limited cognition of decision makers, who are faced with incomplete information about the present and uncertainty regarding the future in an ever-changing economy. Habitual consumer behaviour contributes to inertia in market demand for mature products, while resistance to new products implies gradual diffusion along meso-trajectories shaped by market institutions and social networks.

Chapter 3 discusses pricing behaviour by firms. As with consumers, incomplete information and uncertainty lead to behaviour based on rules and routines, rather than optimisation. In addition, large firms face the task of coordinating the activity of substantial numbers of individuals with highly specialised knowledge, which makes establishing internal and external connections a primary focus of firm strategic planning. Heterogeneity among firms is highlighted, with differences in firm size, scope, organisation, and orientation discussed.

Heterogeneity in firm characteristics combines with heterogeneity in their historical experiences to lead to heterogeneity in pricing behaviour. Price-taker behaviour tends to occur when firms are small relative to the markets in which they operate, especially with organised trading exchanges. Price-maker behaviour occurs when firms have limited numbers of close competitors due to either large size relative to their markets or the differentiation of their products. Price leadership, which is useful for coordinating price behaviour among heterogeneous firms, involves aspects of price-taking behaviour (for price followers) and price-making behaviour (for price leaders). Pricing routines change with industry development, as exemplified by the pricing trajectories of novel products.

Chapter 4 explains the use of the concepts of coordinating price and market order to replace the mainstream concepts of equilibrium price and quantity. Equilibrium is inconsistent with the potential for change from within, which is an essential element of evolutionary analysis. Market order is achieved when bids from potential buyers are closely balanced with offers from potential sellers. In a market with undifferentiated products and numerous buyers and sellers, perfect market order is achieved with all transactions completed that involve bids equal or above the coordinating price and offers equal or below the coordinating price. Administered prices are coordinating prices for markets with heterogeneous products or small numbers of suppliers, with firms holding excess capacity to achieve market order.

Coordinating prices change when external influences cause variations in the bids or offers, much as shifts in supply and demand change the equilibrium price in mainstream price theory. However, the coordinating price also changes due to endogenous processes, such as the process of differential firm growth analysed in Chapter 5. As low-cost firms expand their production capacity relative to their high-cost rivals, more product is offered at low prices. With a homogenous product and price-taker firms, the coordinating price declines as shown in Figure 3. With differentiated products and administered prices, the share-weighted averages across firms of unit cost and prices decline as indicated in (5.21) and (5.19), respectively.

Analysis of price determination in Chapters 4 and 5 demonstrates the dual role of prices and markets in the evolutionary process of organisation and transformation in capitalist economies. Establishing market order is an example of the organising role of prices, while differential firm growth is an example of the transformational role of prices. Prices provide information to buyers and sellers to complete mutually beneficial transactions, while prices combine with heterogeneous costs to finance differential firm growth leading to change in industry structure. Prices contribute to both the self-ordering of capitalist economies and to their self-transformation.

The transformational role of prices also involves providing information to entrepreneurs to identify profitable opportunities for innovation. Heterogeneous expectations or imaginings of the future are an essential driver of endogenous change. Reliable information on current and future prices enhances the confidence in calculating the potential profitability of innovations for both entrepreneurs and their prospective financiers. Realised profitability is the outcome of the market testing of divergent expectations and imaginings. No one knows the future and assuming otherwise is a nonsense, as is assuming that expectations are correct on average.

Chapter 6 discusses fuzzy boundaries for firms and industries. Restless knowledge presents opportunities for innovation both within and outside established firms. Entrants with new processes and products enter established industries or create new industries, provided they have access to finance. Boundaries of firms and industries are changeable when firms exploit innovations in multiple markets, or when diversification offers a strategy for firm growth while avoiding aggressive competition in existing markets. Dynamic capabilities enhance the ability of firms to move into new markets and industries, with mega-firms institutionalising the development of dynamic capabilities to achieve sustainable competitive advantage through self-transformation of the firm.

Chapter 6 also discusses inter-industry linkages in prices from an evolutionary perspective. Rather than assuming a single best-practice technique in each industry, evolutionary analysis incorporates the impact of firm heterogeneity. A hypothetical firm with industry-average values of input requirements, price-cost margins, and prices is used to represent the industry. Creative destruction through differential firm growth leads to declines in unit cost, price-cost margin, and price for the representative firm, even without further technological change at any firm in the industry.

Price decreases spread through the economy as input costs decrease for firms in downstream industries. Importantly, the whole process of price adjustment is an example of evolutionary change. As the process reduces heterogeneity, the rate of change slows. However, the introduction of novelty through entrepreneurial action based on restless knowledge, imagination and creativity ensures continued development from within. Long-run equilibrium is a nonsense. There is a journey, but the destination keeps changing before it is reached.

The micro-meso-macro methodology of evolutionary price theory features in Chapter 7, which discusses movements in economic aggregates in a developing economy. Schumpeter argues waves of innovation have a macro impact, generating cycles in prices of the means of production impacting firms throughout the economy. This adds to movements in prices at the firm level due to innovation, and to movements in prices at the industry level due to creative destruction. The structure of production and the pattern of consumer expenditure change, leading to obsolescence of capital equipment and labour skills. Aggregate price indices imperfectly capture changes in the price level due to changing relative prices, invalidating meaningful comparisons over long periods for aggregate output, the real wage, and the material standard of living. Evolutionary price theory is a theory of the heterogeneous movement in prices, not a theory of stationary general equilibrium or steady-state growth.

8.2 Areas for Further Research

Analyses in previous chapters identify a variety of mechanisms applying to price determination in an evolving economy. Identifying gaps that remain in the analysis is aided by classifying the mechanisms according to two distinguishing characteristics. Table 1 depicts the classification. One distinguishing characteristic is whether firms are price takers or price makers, while the second is whether the market in which they operate is orderly or disrupted by differential firm growth.

Each of the mechanisms of price determination in Table 1 is analysed separately. However, they are all part of a process of economic development that includes transitions between mechanisms. Analysis of the transitions is a proper part of a comprehensive evolutionary price theory.

Creative destruction as analysed in Section 5.2.3 is an example of a transition process between categories of pricing mechanisms. With differential firm growth the market shares of relatively high-cost firms decline while low-cost firms become dominant. As the share-weighted variance in unit cost falls, the driving force of disruption of the market eases.

Innovation results in disruption of orderly markets. Section 3.2.4 discusses factors affecting firm orientation towards innovation. Section 5.1 argues innovation throughout the economy is encouraged by reliable price information coming from orderly markets. All this is sketchy. Research drawing on case studies of innovation is worthwhile for deepening the theory of the role of prices in the transition from orderly to disrupted markets.

Chapter 3 discusses the behaviour of price-taker and price-maker firms without explaining how some firms come to be price takers, while others are price makers. Implicitly, the firm’s organisational structure, history, and institutional setting are influences. The relationship between firms also matters, as is clear in the case of price leadership. Explicitly theorising the process determining whether a firm is a price taker or a price maker is a topic worthy of further research.

Table 1 shows price determination in markets with price-taker firms is analysed by balancing supply and demand. A related assumption of pricing to balance the growth of supply with the expected growth of demand is applied to disrupted markets with price-maker firms. These are simplifying, but restrictive, assumptions for thinking through the implications of firm behaviour for market outcomes. Dropping assumed balance in favour of allowing for excess demand or unutilised capacity, as occurs with the analysis price-maker firms in orderly markets, is a worthwhile step towards generalising the evolutionary theory of price determination.

A weakness in the analysis of price determination in markets with differential firm growth in Chapter 5 is the assumption that the pricing routines of firms are unchanged over time. Firm pricing routines coevolve with market structure. In Section 5.3, integrating analysis of changes in firm pricing routines and the evolution of industry structure is suggested as a useful direction for future research in evolutionary price theory. The work of business historians and management scholars provides a fertile ground for appreciative theorising on this topic.

Hints about integrating analysis of changes in pricing routines with the evolution of industry structure are provided in Section 3.4, where pricing routines change over the origination, adoption, and retention phases for introduction of a novel product. Also relevant is Steindl’s (Reference Steindl1976) post-Keynesian analysis of differential firm growth, which has firm pricing routines becoming less aggressive with increasing industry concentration. Recent works featuring coevolution of pricing routines, consumer behaviour, and market structure include Markey-Towler (Reference Markey-Towler2016) and Almudi et al. (Reference Almudi, Fatas-Villafranca, Palacio and Sanchez-Choliz2020). More research on this topic is worthwhile.

Investment routines are unchanged over time in analysing differential firm growth in Chapter 5, which ignores the potential co-evolution of investment routines and industry structure. In addition to pricing routines becoming less aggressive with rising industry concentration, Steindl (Reference Steindl1976) has firms reducing the amount of profit invested in capacity expansion. A useful direction for further research is deepening the links among investment routines, pricing routines, and the evolution of industry structure. Bloch and Kriesler (Reference Bloch and Kriesler2023) review post-Keynesian analysis on linking pricing to the investment decision.

Metcalfe (Reference Metcalfe and Dopfer2025) suggests a need for extending analysis of pricing with differential firm growth beyond an unchanging group of firms, whose expansion is proportional to their profits. New entrants are a source of imitators to accelerate the diffusion of innovations through external financing of investment in productive capacity. Firms diversifying from related industries are a further source of expansion through imitation. Financing of innovation and its diffusion using external sources, including firms diversifying from other industries, is discussed in Minsky (Reference Minsky, Heertje and Perlman1990), Knell (Reference Knell2015) and Callegari (Reference Callegari2018). Further developing the role of entry and external finance in the analysis of differential firm growth is a prime topic for further research.

Nelson (Reference Nelson2020) observes studies of productivity growth at the firm and industry levels show improvements in productivity within firms account for a larger proportion of increases in industry average productivity than does the selection of firms through differential firm growth. Learning and competitive pressure are important drivers of productivity improvement at the firm level (see Downie (Reference Downie1958) for an early discussion). There is also a bias in the measurement of productivity growth at the firm level. Relative prices of high-cost firms rise with differential firm growth in the analysis of Section 5.2.2. A rising relative price is measured as a rise in real output when the value of output of a high-cost firm is deflated by the share-weighted price for the industry. A topic worthy of further research is the implications of evolutionary price theory for the measurement of productivity growth at the micro, meso, and macro levels.

The scope of this Element is limited. Discussion of applications to empirical research is restricted to a few peripheral comments. How evolutionary price theory might guide economic policy is not discussed. Developing the implications of evolutionary price theory for empirical research and economic policy are prospective topics for further research.

8.3 Final Observations

Evolutionary price theory provides a framework for analysing the dual role of prices in capitalist economies. Prices coordinate mutually beneficial exchanges between buyers and sellers. They also provide information to entrepreneurs to evaluate the profitability of potential innovations arising from human creativity and restless knowledge. Market capitalism is self-ordering and self-transforming, with prices playing an essential role.

Evolutionary price theory abandons the pursuit of a universal science of human economic behaviour that characterises mainstream economics, instead incorporating history and institutions into the analysis of economic activity. An open-systems ontology is employed in place of the closed-system ontology of the mainstream. Furthermore, a micro-meso-macro methodology is used in place of the methodological individualism used in mainstream economics.

Evolutionary price theory has consumers and firms operating according to habits, rules, and routines, rather than assuming optimising behaviour based on the Olympian rationality of mainstream economics. These habits, rules, and routines are reasoned, but imperfect, responses to limited cognition, incomplete knowledge and connections, and an uncertain, indeed unknowable, future. They change with learning and imagination, so they are historically and institutionally specific, co-evolving with institutions and society.

Learning improves the operations of consumers and firms, while entrepreneurial action adds novelty to the market environment. Markets provide a testing ground to sort the wheat from the chaff. Winners expand relative to the losers, increasing the average fitness of the population. However, sorting using the market mechanism with creative destruction comes at a cost to society in terms of obsolete capital equipment and skills. Just as capital is created when bankers extend newly created credit to entrepreneurs and the market validates this capital by profits to the entrepreneur, capital is destroyed when the entrepreneurs drop prices to aggressively to expand their market share. Mainstream economics is silent on such dynamics because all firms are assumed to be identical and operated optimally, which rules out creative destruction.

Evolutionary price theory treats competition as a process unlike the structure-conduct-performance approach of mainstream economics. In the differential firm growth analysis of Chapter 5, prices fall with rising industry concentration as market share shifts from high-cost firms to low-cost firms. Innovation leads to new products and, eventually, to new industries along the meso-trajectory of origination, adoption, and retention, with prices declining as firms adopt more aggressive pricing routines to expand their market.

Schumpeter’s ([Reference Schumpeter1950] Reference Schumpeter1976, p. 85) view of the competitive process remains relevant, ‘it becomes a matter of comparative indifference whether competition in the ordinary sense functions more or less promptly; the powerful lever that in the long run expands output and brings down prices is in any case made of other stuff.’ The competition that counts comes from new products, new production processes, and new ways of organising distribution, witness mobile versus landline telephony, electricity from solar panels and wind turbines versus coal-fired generating plants, and ecommerce versus bricks-and-mortar retailing.

Evolutionary price theory analyses the determination of prices in an economy that is growing through innovation and structural change. It is a theory of the movement of prices through time rather than a theory of the structure of prices in the short or long run as in mainstream economics. The process occurs unevenly over time, is irreversible and path dependent. The pattern and content of the economic order is continually transformed, creating worlds we then lose.Footnote ³⁷

The heterogeneity of firms and consequent price dynamics are at centre stage. Some industries are directly impacted, while other industries face changes in the prices of inputs and the prices of competing products. The process is locally degenerative, with shrinking market shares of unfit firms consuming the heterogeneity that drives differential firm growth. However, entrepreneurial actions based on restless knowledge, creativity, and imagination ensure that variety is replenished. Development from within continues indefinitely. Evolutionary price theory provides the framework for analysing how prices guide and are impacted by this process.