The exploration of industry dynamics | Carlsbergfondet
Til projektoversigt

The exploration of industry dynamics

The Carlsberg Foundation's 'Semper Ardens' Research Project

Do robots, disruption and globalisation really equal doomsday scenarios of mass unemployment and growing inequality? Or do we simply not know enough about the forces and adjustments that new technologies and international commerce trigger in our market economies? By using register data, research can revise our understanding of industry dynamics.

By Professor Philipp Schröder, PhD, Director, FIND Research Centre for Firms and Industry Dynamics, Department of Economics and Business Economics, Aarhus BSS, Aarhus University

In 2016, many researchers - particularly economists and political scientists - had something of a wake-up call when Donald Trump won a surprise victory in the US presidential election with an election programme based on “America First” and economic protectionism. A few months before, the UK had voted to leave the EU.

Drivers of Economic Growth:

Analyses from the Maddison project database (hosted at University of Groningen, NL) suggest that global average per capita income (and accordingly living standards) has multiplied by factor ten in the past 200 years. Economists link this surge to the international division of labour and technological progress (industrial revolutions)

Now on the brink of the fourth industrial revolution the global division of labour based on free trade and technological progress is viewed with mistrust. However, the sober fact is that new international markets and new technologies have been the central drivers of economic growth.  

Shocks to the system

So, why the popular backlash to these forces today? Globalisation and technological progress always operate through the individual business decisions of firms. 

Firms develop new technologies and access new international markets in search of profit. In doing so they replace old technologies and pressure incumbent firms, which might have to close and shed jobs in response to new technologies or foreign competition. 

Such closures are the ingredients for the doomsday scenarios peddled by political actors across the globe.  By understanding the detailed mechanisms - the so-called industry dynamics - of these immense changes in our economic systems, we can gain new and important insight into the workings of market forces. Eventually, we will understand by which channels technological progress and globalisation affect firms and individuals, and which policies can remedy the downsides while harvesting the gains.

Picture 1: While economic science conclusively has established the benefits of international trade and technological progress for economic well-being, an increasing number of countries and regions see a popular backlash towards globalisation and technology. Photo of container ship: Cuttersnap/Unsplash.

The recipe for capitalism

Free market forces represent a form of economic Darwinism where companies that are not up to speed are driven out of business by companies that are. However, while weaker companies are forced to shut down and lay off employees, the strong companies grow and hire more people. In this process of creative destruction new companies and new types of jobs emerge. It is the total picture that counts, and here new research may contribute to drawing a more nuanced picture.

Register data on the entire population of companies


Picture 2: Mathematical and statistical methods applied to large-scale administrative register data allow researchers to model the industry dynamics of market economies. Photo by Markus Spiske on Unsplash.

Research Methodology:

Research at the FIND Centre utilises large-scale administrative register data of all firms in a given country, their activities and employees. Advanced mathematical and statistical methods allow researchers to infer relationships between aggregate outcomes, such as economic growth and specific events such as new technologies or export decisions

Despite the apparent simplicity of the industry dynamics concept, namely entry and exit of firms or changing market shares, research faces significant obstacles. Data availability is often limited and theoretical formulations of dynamics swiftly increase in their mathematical complexity. Backed by a large-scale grant from the Carlsberg Foundation, the research project FIND (Firms and INdustry Dynamics) at Aarhus BSS, Aarhus University, is now tackling these obstacles and increasing our understanding of industry dynamics.

By combining extensive and detailed register data at the individual and company levels, we are able to carry out novel empirical analyses of the industrial shifts that we witness. By incorporating these movements into macroeconomic models we will, among other things, be able to conduct policy experiments and predict the economic consequences of various initiatives, say the potential effects of a robot tax as a remedy to technological disruption. 

Thus, ultimately, a deep understanding of industry dynamics ensures that political considerations can be founded on evidence-based research.

Export customers that never return and robots that create jobs

Analysing market forces at this granular level gives new insights. For examples: the FIND project has established that companies’ export patterns are far more nuanced than previously believed.

By combining customs data for 221,000 Danish export sales with company data, we have made the surprising discovery that approximately one third of all export sales are one-off occurrences - i.e. export sales that the company in question is unable to repeat. Our research is the first to reveal this substantial presence of passive exports in the total exports of Denmark.

In a different line of research based on Spanish register data the FIND project has analysed the effects of robots on employment. Contrary to what one might expect,

Contrary to what one might expect, we find that companies that introduce robots in manufacturing end up hiring more employees. While in companies that do not introduce robots, the head count ends up decreasing.

Picture 3: Data plot of employment in robot adopting firms versus non-adopting firms. Evolution of average firm employment (measured by the number of workers) in a balanced sample of Spanish firms 1990-2016. Computations based on ESEE register data by Michael Koch (FIND Research Centre, Aarhus University), Ilya Manuylov (FIND Research Centre, Aarhus University) and Marcel Smolka (Europa-University Flensburg).

To sum up: viewing the market economy through the lens of industry dynamics gives new insights and allows us to revise our understanding of aggregate phenomena, such as inequality. Eventually, this understanding will help to inform policy decisions.  Maybe most importantly, industry dynamics continue to pose many new questions.

Academic References:

Geishecker, Ingo, Philipp Schröder and Allan Sørensen (2019). “One‐off export events”, Canadian Journal of Economics 52(1), pp 93-131.

Schröder, Philipp and Allan Sørensen (2012). “Firm exit, technological progress and trade”, European Economic Review 56( 3), pp. 579– 591.

Acemoglu, Daron and Pascual Restrepo (2018). “The race between man and machine: Implications of technology for growth, factor shares, and employment”, American Economic Review 108(6), pp.1488-1542.

Bernard, Andrew, Teresa Fort, Valerie Smeets and Frederic Warzynski (2020). “Heterogeneous Globalization: Offshoring and Reorganization”, NBER Working Paper No. 26854.

Hopenhayn, Hugo (1992). ”Entry, Exit, and firm Dynamics in Long Run Equilibrium”, Econometrica 60 (5), pp. 1127-1150.