The complexity of the intangible digital economy: an agent-based model

https://doi.org/10.1016/j.jbusres.2020.03.041Get rights and content

Highlights

  • Agent-based macro model to address micro-macro effects of the digital transformation.

  • Digital-asset competitiveness leads to higher productivity.

  • Converse concentration effect: concentrated digital markets with low average prices.

  • Long-term technological unemployment due to increasing digital assets productivity.

Abstract

During the last decades, we have witnessed a strong development of intangible digital technologies. Software, artificial intelligence and algorithms are increasingly affecting both production systems and our lives; economists have started to figure out the long-run complex economic implications of this new technological wave. In this paper, we address this question through the agent-based modelling approach. In particular, we enrich the macroeconomic model Eurace with the concept of intangible digital technology and investigate its effects both at the micro and macro level. Results show the emergence of the relevant stylized facts observed in the business domain, such as increasing returns, winner-take-most phenomena and market lock-in. At the macro level, our main finding is an increasing unemployment level, since the sizeable decrease of the employment rate in the mass-production system, provided by the higher productivity of digital assets, is usually not counterbalanced by the new jobs created in the digital sector.

Introduction

During the course of history, several technological discoveries have influenced the lives of human beings. In this paper we focus our attention on the impact of digital transformation on production systems and, as a consequence, we evaluate the potential variation of the employment rate in the long term. According to Brynjolfsson and McAfee (2014), we are facing “The Second Machine Age” that is revolutionising our world. In particular, the authors argue that probably one of the most important technological discoveries was the steam engine, created by John Watt in the second half of the eighteenth century, which allowed production of a huge amount of mechanical energy. After that, there were further technological developments that affected our production systems and, thanks to electronics and information technology, in the second half of the twentieth century assembly lines were largely automated. Nowadays, we are facing a new technological wave, indeed, digital technologies have been the subject of intense improvement and the possible consequences of this productivity enhancement are currently being debated among economists.

The potential effects of technological transitions on the labour market have been the subject of a long debate among economists since the first industrial revolution. The potential outcomes deriving from technological progress have been distinguished between: short-term disruption and long-term benefits, see Mokyr, Vickers, and Ziebarth (2015). In fact, according to the “Compensation Theory”, in the long-term, compensation mechanisms counterbalance the unemployment created by technological progress, see Vivarelli (2014). Along this line, technological unemployment is only temporary: the economy experiences a structural change rather than the so-called “end of work”, Vermeulen, Kesselhut, Pyka, and Saviotti (2018).

However, the nature of new digital technologies is different compared to machines deriving from the steam engine and traditional automation. The substantial difference between digital technologies and traditional industrial automation is that while the latter helps human beings to overcome the limits linked to physical force, thanks to the former we can surmount the limits imposed by our mind. Moreover, several economists and technologists argue that artificial intelligence, thanks to significant improvements in computation, could become self-improving causing a technological singularity, see Good, 1966, Nordhaus, 2015, Aghion et al., 2017.

According to Acemoglu and Restrepo, 2017, Acemoglu and Restrepo, 2018a, Acemoglu and Restrepo, 2018b, Acemoglu and Restrepo, 2018c, Acemoglu and Restrepo, 2018d, AI and robotics, as automation, replace human beings in jobs that they previously performed, creating a “displacement effect” and this destruction of job places could only be effectively countervailed by the creation of new labour-intensive tasks. Moreover, empirical evidence shows a labour market polarisation: whereas technology until the end of the XX century impacted principally on workplaces occupied by “blue-collar” workers, probably these kinds of digital instruments will mainly affect the so called “white-collar” workers performing jobs which require routine manual and cognitive skills, see Goos and Manning (2007).

Furthermore, it is really interesting to notice how the business dynamics related to companies which develop and produce digital technologies are completely different compared to the economic dynamics that characterised mass-productions. As a matter of fact, Arthur, 1989, Arthur, 1990, Arthur, 1994, Arthur, 1996 distinguishes between two different worlds: a mass-production world, characterised by diminishing returns, in which products are heavy on resources and light on knowledge and a knowledge-based world that, on the contrary, is characterised by increasing returns. In this particular reality, which regards high-tech producers, products require a deep know-how and scarce quantity of resources; in other words, these companies have high R&D fixed costs compared to their variable production costs. Furthermore, according to Arthur, a world ruled by increasing returns presents several other characteristics such as network effects, path dependence, market instability, unexpectedness, winner-take-all and technological lock-in. These features are being studied in a field called Complexity Economics which, unlike standard economic theory, emphasises interaction among economic agents through an out-of-equilibrium approach, see Elsner et al., 2014, Arthur, 1999, Arthur, 2014, Fontana, 2010.

Agent-based modelling represents an appropriate approach in order to address these aspects, see Gallegati, 2018, North and Macal, 2007, Hommes and LeBaron, 2018. Out-of-equilibrium dynamics, complex interactions among economic agents and heterogeneity are three important features that can be encompassed by agent-based modelling. Since the AI advent can be framed as a transition phase in the history of technological progress, an out-of-equilibrium approach, such as the agent-based one, can be an effective way to represent this structural and productive transformation. Furthermore, by capturing heterogeneity between economic agents we can distinguish between different types of productive capital: hard capital and intangible or digital capital. The need for heterogeneity to study the potential effect of a digital transformation is also reflected by the labour force: workers are heterogeneous and they differ in skills. Finally, interactions drive several features of the “increasing returns” world, such as for example network effects, lock-in and winner-take-most-phenomena.

In this paper, we enrich a pre-existing large-scale macroeconomics model, called Eurace (see Mazzocchetti et al., 2018, Ponta et al., 2018, Raberto et al., 2012, Teglio et al., 2012) to tackle our research questions. The concept of innovation has already been investigated by means of agent based models (see e.g. Caiani et al., 2019, Dawid and Reimann, 2011, Dosi et al., 2010, Fanti, 2018, Pyka et al., 2010, Vermeulen and Pyka, 2014, Vermeulen and Pyka, 2018), and also the Eurace model has been endowed with the concept of innovation, see Dawid et al., 2008, Dawid and Gemkow, 2014, Dawid et al., 2014, Dawid et al., 2018, Dawid et al., 2019. However, we focus here on innovation from the perspective of productivity increases due to intangible digital capital goods, not only tangible ones. Software, algorithms, artificial intelligence and their developers are the subject of our study, as we want to link the concept of innovation to the one of “digital revolution”, as described in Brynjolfsson and McAfee (2011). The addition of digital technologies in the Eurace model mimics the advent of Industry 4.0, according to which not only are the production processes automated, but also decisions start to be subject to automation technology, see Kang et al., 2016, Parrott and Lane, 2017, Cotteleer and Sniderman, 2017. From a macro perspective, the research work tries to address and evaluate the potential effect of a digital transformation on the economic system. Furthermore, at a micro level, our analysis aims to study the main business dynamics characterising digital technology producers. In this respect, the novelty of our contribution concerns the introduction of a new type of capital producer within a large-scale macroeconomic agent-based model: the intangible or digital assets developer.

The introduction of this new kind of firm, which belongs to the “increasing returns world”, turns out to be crucial in order to better understand and investigate the economic implication of digital technologies on business, both from a macro and micro point of view. In fact, being a bottom up approach, agent-based modelling gives us the opportunity to study not only the macroeconomic trend of the system but also the sectorial behaviours.

The new Eurace model features regarding the production of digital intangible technologies are presented in Section 2. Section 3 shows our preliminary computational results. The conclusion and remarks are provided in Section 4.

Section snippets

Outline of the Eurace model

A description of the baseline version of the Eurace model that has been used in this paper can be found in Teglio, Mazzocchetti, Ponta, Raberto, and Cincotti (2019), while Petrovic, Ozel, Teglio, Raberto, and Cincotti (2017) explains the model in more detail.1 In this section we recall the basic features of the model that can be useful for

Design of experiments

The new features of the model allow us to analyse different scenarios. In particular, we consider two digital asset pricing scenarios. In the first one, named “collusive pricing”, DADs sell their licences at the same price, determined as a fixed share of the nominal wage. In the second one, henceforth “competitive pricing” scenario, we endowed the firm with the possibility of independently raising or decreasing their licence prices; the choice between these two options depends on the market

Conclusion

The computational results presented in the paper are able to capture the essence related to the new digital technologies world and the stylised facts that characterise the existing literature. Furthermore, the economic dynamics that emerged from the simulations show interesting properties both at the micro and at the macro levels. The existing differences between competitive and collusive pricing point out very interesting aspects. Both cases lead to the success of a company with respect to

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

University of Genova is acknowledged for providing a PhD grant.

Filippo Bertani received his B.S. degree in Mechanical Engineering in 2015 and his M.S. degree in Mechanical Engineering – Design and Production in 2017 from the University of Genoa, Italy. Since November 2017 he has been doing a Ph.D. course in Mechanical Engineering, Energetics, Management and Transportation with a curriculum in Economy and Management at the University of Genoa joining the Research Center on Organization Science, Economics and Management (DOGE). He participated at several

References (66)

  • D. Acemoglu et al.

    Low-skill and high-skill automation

    Journal of Human Capital

    (2018)
  • D. Acemoglu et al.

    Modeling automation

    AEA Papers and Proceedings

    (2018)
  • P. Aghion et al.

    Artificial intelligence and economic growth. Working Paper 23928

    (2017)
  • W.B. Arthur

    Competing technologies, increasing returns, and lock-in by historical events

    Economic Journal

    (1989)
  • W.B. Arthur

    Positive feedbacks in the economy

    Scientific American

    (1990)
  • W.B. Arthur

    Increasing returns and path dependence in the economy

    (1994)
  • W.B. Arthur

    Increasing returns and the new world of business

    Harvard Business Review

    (1996)
  • W.B. Arthur

    Complexity and the economy

    Science

    (1999)
  • W.B. Arthur

    Complexity and the economy

    (2014)
  • P. Belleflamme et al.

    Platforms and network effects

    Handbook of Game Theory and Industrial Organization: Applications

    (2018)
  • E. Brynjolfsson et al.

    Race against the machine: How the digital revolution is accelerating innovation, driving productivity, and irreversibly transforming employment and the economy

    (2011)
  • E. Brynjolfsson et al.

    The second machine age: Work, progress, and prosperity in a time of brilliant technologies

    (2014)
  • A. Caiani et al.

    Does inequality hamper innovation and growth? An AB-SFC analysis

    Journal of Evolutionary Economics

    (2019)
  • C.D. Carroll

    A theory of the consumption function, with and without liquidity constraints

    Journal of Economic Perspectives

    (2001)
  • M.M. Cotteleer et al.

    Forces of change: Industry 4.0

    (2017)
  • H. Dawid et al.

    How do social networks contribute to wage inequality? Insights from an agent-based analysis

    Industrial and Corporate Change

    (2014)
  • H. Dawid et al.

    Skills, innovation, and growth: An agent-based policy analysis

    Journal of Economics and Statistics

    (2008)
  • H. Dawid et al.

    Macroeconomics with heterogeneous agent models: fostering transparency, reproducibility and replication

    Journal of Evolutionary Economics

    (2019)
  • H. Dawid et al.

    Diversification: a road to inefficiency in product innovations?

    Journal of Evolutionary Economics

    (May 2011)
  • U. Doraszelski et al.

    Measuring the bias of technological change

    Journal of Political Economy

    (2018)
  • W. Elsner et al.

    The microeconomics of complex economies

    (2014)
  • S. Fabiani et al.

    What firms’ surveys tell us about price-setting behavior in the Euro area

    International Journal of Central Banking

    (2006)
  • Fanti, L. (2018). An AB-SFC Model of Induced Technical Change along Classical and Keynesian Lines. Mpra paper. Germany:...
  • Cited by (55)

    • To copatent or not to copatent: An agent-based model for firms facing this dilemma

      2023, European Journal of Operational Research
      Citation Excerpt :

      Table 1 provides information on the domain of the symbols and their range. The computational results are presented in the form of boxplots, where each boxplot exhibits the distribution of the time averages of firm performance variables over a twenty-year time interval concerning the 150 simulations characterized by different seeds (Bertani, Ponta, Raberto, Teglio, & Cincotti, 2021). The analysis aims to study the firms’ behaviour at a micro-level in order to verify the impact of copatenting on firm performance, considering both economic and innovation performance.

    View all citing articles on Scopus

    Filippo Bertani received his B.S. degree in Mechanical Engineering in 2015 and his M.S. degree in Mechanical Engineering – Design and Production in 2017 from the University of Genoa, Italy. Since November 2017 he has been doing a Ph.D. course in Mechanical Engineering, Energetics, Management and Transportation with a curriculum in Economy and Management at the University of Genoa joining the Research Center on Organization Science, Economics and Management (DOGE). He participated at several international conferences and workshops and his main research topics are related to agent-based modelling, macroeconomics, finance, innovation, sustainability and industrial symbiosis. In particular, he is working on digital technologies innovation and its effect on the labour market. He is member of the European Association for Evolutionary Political Economy (EAEPE).

    Linda Ponta is assistant professor in Management Engineering at LIUC - Cattaneo University where she teaches Financial Engineering, Business Process Management and Control and Business Decision Making: Models and Tools. She received the M.S. degree in electronics engineering (summa cum laude) and the Ph.D. degree in electronics and computer engineering from the University of Genoa, Genoa, Italy, in 2004 and 2008, respectively. She joined the CINEF Group, Center for Polymer Studies (supervised by Professor H. E. Stanley), Boston University, and the Politecnico di Torino, starting work in financial markets. Linda won the “Bruno de Finetti 2006” prize awarded by Accademia Nazionale dei Lincei. Her primary research interests are economics and financial engineering, nonlinear circuits, networks, econophysics, and innovation.

    Marco Raberto is Associate Professor of Management Engineering at the University of Genoa, Italy. He holds a MSc degree in Physics (1999) and a PhD in Computer Science Engineering (2003). His main research interests focus on financial stability, green finance and digital transformation under the agent-based modelling and complexity approaches. He is author of more than 50 peer-reviewed scientific publications and organized several workshops and conferences on these subjects, such as EAEPE 2015. He participated at international research project on agent-based modelling and simulation in economics and finance. In particular, he has been one of the leading researchers in project EU-FP6 Eurace (2006–09) and in project EU-FP7 Symphony (2013–16). He has been the Principal Investigator of project Iceace (2011–13) funded by the Icelandic Centre for Research (Rannis). Marco Raberto is the scientific development plan officer and research area coordinator of the European Association for Evolutionary Political Economy (EAEPE) since 2014.

    Andrea Teglio is Associate Professor at the Department of Economics of Ca’ Foscari University of Venice, Italy. He holds a PhD in Economics (Universidad Jaume I de Castellon, 2011) and a PhD in Electronic and Computer Science Engineering (University of Genova, 2004). His main research interests are macroeconomics, agent-based modelling, complexity science, nonlinear dynamics and chaos. He is author of several scientific articles published in peer-reviewed international journals. He has been principal investigator in several research projects, among which the EU-FP7 Symphony (2013–16), and he has organized conferences and workshops on his favourite topics (e.g. WEHIA 2016, NAIM 2012, Artificial Economics 2012). He is associate editor of the Review of Evolutionary Political Economy and of the International Journal of Microsimulation.

    Silvano Cincotti (ME 1990, PhD 1994) is professor of Economics and Management at the University of Genoa. He is head of DOGE and CINEF research groups, dean of the IANUA School of Advanced Studies at the University of Genoa and CEO of the IANUA Consortium. He is editor of the REPE and associate editor of the JEIC and EIER. His current research topics includes agent-based macroeconomics and corporate strategy and social responsibility. He is the author of more than 180 scientific articles, member of numerous scientific committees and invited speaker to several international conferences. His scientific activities have been carried on within more than 50 national and international project grants and he was coordinator of EU FP6 IST-FET EURACE and EU FP7 GSS SYMPHONY, partner of EU NEST-PATHFINDER COMPLEXMARKET and steering board member of the Italian HUB of the EU-FET Flagship FuturICT. He is member of AiIG, ESHIA and EAEPE.

    View full text