Synonyms

Nature of technology; Technology; Technological innovation; Technological evolution; Technological change; Technological progress; Technological advances

Definition

Hypothesis refers to a supposition put forward in a provisional manner and in need of further epistemic and empirical support. Technology analysis explains the relationships underlying the source, evolution, and diffusion of technology for technological, economic and social change. Technology analysis considers technology as a complex system that evolves with incremental and radical innovations to satisfy needs, achieve goals, and/or solve problems of adopters to take advantage of important opportunities or to cope with consequential environmental threats.

Introduction

Technology has an important role for competitive advantage of firms and nations, for industrial and economic change in society (Arthur 2009; Hosler 1994; Sahal 1981). Technology can be defined as a complex system, composed of more than one entity or subsystem and a relationship that holds between each entity and at least one other entity in the system (Coccia 2019a, b, d). Technology satisfies needs, and sustains the achievement of goals, and solution of problems of adopters to take advantage of important opportunities or to cope with consequential environmental threats (Coccia 2019a, b). Technology is driven by inventions of new things and new ways of doing things that are transformed into usable innovations in markets for supporting adaptation and/or survival of adopters in highly differentiated and volatile environments (Coccia 2018b, 2019b, c). Technology generates technological, economic and social change by (Coccia 2005b, 2006b, 2016a) incremental innovations (progressive modifications of existing products and/or processes), radical innovations (drastic changes of existing products/processes, or creation of new products/processes to satisfy needs or solve problems in society), technological systems (clusters of innovations that are technically and economically interrelated, e.g., nanotechnology; cf., Coccia and Wang 2015), and technological revolutions (pervasive changes in technology affecting many branches of the economy, such as general purpose technologies of information and communication technologies having a technological dynamism and a pervasive use in a wide range of sectors; cf., Coccia 2015a, 2017a, b, 2020).

Technology analysis can be based on “multiple working hypotheses” (Chamberlin 1897) to explain the relationships underlying the source, evolution, and diffusion of technologies and provide theoretical, empirical, and policy implications. The method of multiple working hypotheses (in short, MWHs) involves the development, prior to research, of several hypotheses that might explain the relationships of specific phenomenon, which is likely due to several causes, not just one (Chamberlin 1897). All suggested hypotheses are considered for scientific investigation, including the possibility that none of them are correct and that some new explanations may emerge (Coccia and Benati 2018; Heidelberger and Schiemann 2009).

MWHs for technology analysis can be schematically synthetized as follows (Fig. 1):

  • Traditional MWHs are demand of technology, induced innovation, learning processes, specialization via scale, disadvantage of beginning, path-dependent processes, competitive substitution, and predator-prey relationships between technologies.

  • New MWHs are based on multimode relationships between technologies, such as the hypothesis of killer technologies and technological parasitism.

Fig. 1
figure 1

Multiple working hypotheses (MWHs) for technological analyses

Full size image

Multiple Working Hypotheses for Technological Analyses

Traditional and new hypotheses for technology analysis provide a method of inquiry to explain how technology originates and evolves in industrial dynamics to generate technical progress in society (cf., Fig. 1).

Traditional Multiple Working Hypotheses for Technology Analysis

Hypothesis of Demand for Technology

This hypothesis suggests that the innovative output of an industry varies in a direct relation to the volume of its sales. Schmookler and Brownlee (1962) argue that the relationship between technological innovation and demand is postulated to hold in long and short run. The demand-pull hypothesis has received convincing evidence with the work by Griliches (1957) that in the study of the invention and diffusion of hybrid maize demonstrates the role of demand in determining the timing and location of invention and innovation. Schmookler (1962, 1966), using patent statistics of inventions in different industries (e.g., railroads, agricultural equipment, paper, and petroleum), shows that the demand is more important in stimulating inventive activity than advances in the state of knowledge.

A simple model to analyze the hypothesis of demand for technology, considering for instance, farm tractor technology, is given by:

$$ \log\ {\mathrm{Y}}_{\mathrm{t}}=a+{\beta}_1{{\mathrm{logX}}^{\#}}_{\mathrm{t}}+{\beta}_2\log\ {\mathrm{Y}}_{\mathrm{t}-1} $$

Y is a measure of technical efficiency; X# is gross investment in tractors, e.g., the number of tractors sold each year (in hundreds). Method of ordinary least squares can be applied for estimating the unknown parameters of relationship just mentioned.

Induced-Innovation Hypothesis

Hicks (1932, pp. 124–125) argues that “a change in the relative prices of factors of production is itself a spur to innovation and to inventions of a particular kind directed at economizing the use of a factor which has become relatively expensive.” The microeconomic version of induced innovation was advanced by Ahmad (1966) and elaborated by Binswanger (1974). In the 1970s and 1980s, theoretical and empirical studies by agricultural economists explain source and evolution of technology with induced-innovation hypothesis (Hayami and Ruttan 1970; Binswanger and Ruttan 1978). Olmstead and Rhode (1993, p. 102) argue that Hayami and Ruttan’s induced-innovation hypothesis reveals two distinct variants. The first is change variant, associated with the argument by Hicks (1932): a rise in the relative price in one factor leads to technological innovations sparing that factor. The second is level variant that even at constant relative factor price levels, new technologies are developed and adopted to save relatively expensive factors (cf., Coccia 2016b).

Learning by Doing Hypothesis

This hypothesis suggests that technical progress depends on acquisition of practical experience for the solution of consequential problems, such as for treating mutant cancers (cf., Coccia 2015b, 2016a, c, 2017b, 2020). In particular, learning by doing hypothesis argues that the evolution of technology is governed by a process of cumulative change, rather than by a set of replicative events at work (Coccia 2014b, 2016a). The operationalization of this hypothesis requires a measure of experience, such as cumulated production quantities or cumulated years of production (cf., Sahal 1981, p. 112). A relationship, which investigates the learning by doing hypothesis of technological innovation, is given by:

$$ \log\ {\mathrm{Y}}_{\mathrm{t}}\, =\, a\, +\, {\beta}_1{\mathrm{logX}}_{\mathrm{t}}\, +\, {\beta}_2\log\ {\mathrm{Y}}_{\mathrm{t}-1} $$

Y is a measure of technical efficiency; X is cumulated production quantities.

Learning Via Diffusion Hypothesis of Technology

This perspective suggests that the increased adoption of a technology paves the way for improvements of its characteristics. In this context, the relevant variable of innovation is the cumulated utilization of technology (i.e., capital stock) rather than cumulated production volume. For instance, the successful development of transport technology depends on how well it dovetails with the larger system of its use and improvements in communications network (cf., Sahal 1981, p. 117).

A relationship that analyzes the learning via diffusion hypothesis of technological innovation is given by:

$$ \log\ {\mathrm{Y}}_{\mathrm{t}}=a+{\beta}_1\mathrm{logX}{\ast}_{\mathrm{t}}+{\beta}_2\log\ {\mathrm{Y}}_{\mathrm{t}-1} $$

Y is a measure of technical efficiency; X* is the stock of technology.

Disadvantage of Beginning Hypothesis of Technology

Vital factors of the hypothesis of disadvantage of beginning can be resistance to change, the effect of sunk costs (costs that have been incurrent and cannot be recovered), a new technology that cannot conform to specification of existing plant, infrastructure and/or equipment, etc. (Frankel 1955; Sahal 1981, p. 115). This hypothesis can imply that the younger the age of capital stock, the better are the prospects for technical progress: technological innovation can be limited as capital stock grows and becomes older. The age variable (i.e., oldness) can be measured as a ratio of capital stock to gross investment. A relationship, based on this hypothesis in the case study of farm tractor technology, is given by:

$$ \log\ {\mathrm{Y}}_{\mathrm{t}}\, =\, a\, +\, {\beta}_1\log {\mathrm{X}}_{\mathrm{t}}^{{\prime\prime}}\, +\, {\beta}_2\log\ {\mathrm{Y}}_{\mathrm{t}-1} $$

Y is a measure of technical efficiency; X″ is the ratio of the number of tractors on farms to the number of tractors sold.

Specialization Via Scale Hypothesis of Technology

Technology depends on scale of its utilization, which is associated with factors of the physical nature of technology itself. For instance, technological advances in electricity generation have been made possible by an increase in the scale of electricity transmission network: the reason is that capacity increases with the square of the voltage (Meeks 1972, p. 74). According to this hypothesis, variations of scale affect the course of innovative activity. This approach considers the relevance of scale to innovation processes that is based on systemic nature of technological progress (Sahal 1981, p. 119). A basic variable here is the scale of input utilization. A relationship to analyze this hypothesis in the case study of farm tractor is given by:

$$ \log\ {\mathrm{Y}}_{\mathrm{t}}\, =\, a\, +\, {\beta}_1\log {\mathrm{X}}_{\mathrm{t}}^{\prime}\, +\, {\beta}_2\log\ {\mathrm{Y}}_{\mathrm{t}-1} $$

Y is a measure of technical efficiency; X′ is the average acreage per farm, which is a main indicator of the scale of input utilization.

Path Dependence Hypothesis of Technology

The concept of path dependence of technological innovation was advanced by Arthur (1989, 1994). David (1985, 1993) provides evidence of the path-dependence perspective with historical studies (such as in QWERTY typewriter keyboard, etc.) and shows how lock-in effects can induce the persistence of inefficient structure of keyboard (i.e., adopters of this technology depend on a vendor for products/processes and services or previous technologies, unable to use another vendor or technology without substantial switching costsand barriers). The path-dependence model is due to a sequence of historical events and choices of techniques that may influence the future pathways of technology. However, the concept of technology lock-in for path dependence seems to work only for network of information and communication technologies characterized by increasing returns to scale. Instead, industries with constant or decreasing returns to scale, lock-in effect does not work. In short, this perspective suggests that technical change is a path-dependent process: it evolves from earlier technological trajectories.

Hypothesis of the Competitive Substitution of Technology

Fisher and Pry (1971, p. 75) show that technological evolution consists of substituting a new technology for old one, such as the substitution of coal for wood, hydrocarbons for coal, etc. Fisher and Pry (1971) model the evolution of a new product or process (emerging technology) that substitutes for a prior one (mature technology) in the form of f / (1-f) as a function of time on semilog paper (f is the market share of the emerging product or process versus time).

Predator-Prey Hypothesis

Technologies can generate a predator-prey relationship: one technology enhances the growth rate of the other, but the second inhibits the growth rate of the first (Pistorius and Utterback 1997, p. 74). In fact, a predator-prey relationship can exist between an emerging technology and a mature technology, in particular when emerging technology enters into a niche market of mature technology, reducing market share of established technology. Farrell (1993) used Lotka-Volterra equations to examine a predator-prey relationship between technologies, such as nylon versus rayon tire cords, telephone versus telegraph usage, etc. Overall, then, a predator-prey interaction has an emerging technology in the role of predator and a mature technology as prey. However, one can also visualize a situation where a mature technology is predator and emerging technology is prey (Pistorius and Utterback 1997, p. 78). Utterback et al. (2020) show this type of predator-prey relationship between plywood and oriented strand board technology (OSB) in a specific period of time (OSB is a composite of oriented and layered strands, peeled from widely available smaller trees).

New Multiple Working Hypotheses for Technology Analysis

Hypothesis of Killer Technologies

Killer technology is a radical innovation that with high technical and/or economic performance destroys the usage value of established techniques previously sold and used in markets (Coccia 2017d, e, 2019c). Killer technology can explain the behavior of innovations that generate a destructive creation for technical and industrial change in markets (Coccia 2019c). An example of killer technology is the diffusion of Solvay process that in the 1900s destroys the Leblanc process in the production of soda (Freeman 1974; Rosenberg 1976). To explore the behavior of killer technologies, a simple model shows how killer technologies destroys established technologies, generating technological change in markets. Let a killer technology = Kl (a new radical technology), and let a victim technology = V (established technology), the model is (Coccia 2019c):

$$ \log Kl=\log A+B\log V $$
(1)

B is the coefficient of growth and indicates different patterns of technological evolution:

  • B < 1, whether new technology Kl destroys at a lower relative rate of change the old technology.

  • B = 1, killer technology Kl substitutes victim technology at a proportional rate of change.

  • B > 1, whether killer technology Kl destroys victim technology at greater relative rate of change.

This model (1) of the evolution of killer technology has linear parameters that are estimated with the ordinary least squares method.

Hypothesis of Technological Parasitism

Coccia (2019a, b, d; Coccia and Watts 2020) proposes the theory of technological parasitism to explain the evolution of technology considering a parasite-host relationship between technologies that generates the coevolution of overall complex system of technology. In particular, Coccia (2019d) argues that technologies have a behavior similar to parasites because technologies cannot survive and develop as independent systems per se, but they can function and evolve if and only if they are associated with other technologies, such as audio headphones, wireless speakers, software apps, etc. that function if and only if they are associated with host or master technology of electronic devices, such as smartphone, radio receiver, television, etc. In fact, a parasitic technology P in a host or master technology H is a technology that during its life cycle is able to interact and adapt into the complex system of H, generating coevolutionary processes to satisfy human needs and/or solve problems in society. This behavior of technologies can be generalized with the theorem of not independence of any technology by Coccia (2018a): the long-run behavior and evolution of any technological innovation Ti is not independent from the behavior and evolution of the other technological innovations Tj. The theory of technological parasitism by Coccia (2019d) proposes a model to explain the evolution of technology with a relationship between a host or master technology (H system) and a parasitic technology (P subsystem). The logarithmic form of model is a simple linear relationship (Coccia 2019a, b, d):

$$ \log P\, =\, \log a\, +\, {B}^{\diamond}\;\log\;H\, +\, {\mathrm{u}}_t $$

  • P = evolutionary advances of parasitic technology

  • H = evolutionary advances of host or master technology

  • loga = constant

  • ut = error term

B ⋄ measures the evolution of parasitic technology P compared to host or master technology H. This theory suggests properties of the evolution of technology (Coccia 2018a, 2019a, b, d; Coccia and Watts 2020):

  1. 1.

    The long-run behavior and evolution of any technology depend on behavior and evolution of inter-related technologies (Coccia 2018a, 2019d).

  2. 2.

    Technological host or master with many parasitic technologies generates a rapid evolution of technological host-parasite system. Technological systems with fewer parasitic technologies and a low level of interaction with other technologies improve slowly (Coccia and Watts 2020).

  3. 3.

    Property of mutual benefaction between interactive technologies argues that the interaction between technologies reduces negative effects and favors positive effects directed to an evolution of reciprocal adaptations of technologies to satisfy needs and/or solve problems of adopters in society (Coccia 2019d).

Conclusion

Determinants of technology and technological evolution are due to manifold factors, such as research and development investments, technology transfer, appropriate social structures with consolidated democracy, good economic governance, widespread higher education system, skilled human capital, moderate growth rates of population, purposeful nations with high economic-war potential, etc. (Coccia 2005a, c, 2006a, 2010, 2014a, c, 2015a, c, 2017c). Hence, technology analysis needs a method of inquiry based on multiple working hypotheses for a comprehensive explanation of relations supporting origin, evolution, and diffusion of technology, rather than applying a single hypothesis in isolation that can provide partial explanations and likely misleading results.

Cross-References