The Disruptive Substitution of Human Capital by IT

Master's Thesis, 2017

59 Pages, Grade: 2


Table of content

Table of figures

List of tables

List of abbreviations

1.1 Problem formulation
1.2 Objective of the Master Thesis
1.3 Course of investigation

2 Glossary of terms
2.1 Substitution
2.2 Disruption
2.3 Information Technology
2.4 Unemployment

3 Main part
3.1 Unemployment statistics and trend analysis
3.2 Frey's and Osborne's “The Future of Employment” study
3.3 The OECD's “Risk of Automation for Jobs in OECD Countries” Report
3.4 The WEF's “The Future of Jobs” Report
3.5 Key commonalities and differences

4 Conclusions and future prospects



Table of figures

Figure 1 : The S curve concept

Figure 2: The Second Disruptive Force: Accelerating Technological Change

Figure 3: The global unemployment rate 1991-2014

Figure 4: The unemployment rates 1991-2014 for the world's largest economies and economic unions

Figure 5: Unemployment rates as a function of EGDI and DEI for the world's largest economies

Figure 6: The WEF's 2015-2020 job growth and contraction outlook along nine key job families

Figure 7: Mapping of and correlations between the WEF's employment outlook and Frey's and Osborne's IT substitution probability

Figure 8: The OECD unemployment rate 1991-2014

List of tables

Table 1: ISIC levels and nomenclature of IT-specific sectors

Table 2: Indices for the measurement of digital evolution

Table 3: Correlation between unemployment rate and digital indices

Table 4: Time series of China's, India's and the U.S.'s unemployment rate according to the International Labour Organization, EGDI and DEI

List of abbreviations

illustration not visible in this excerpt


In the more recent history of mankind industrial revolutions have given rise to new technologies in the economy's primary and secondary sector. The higher the degree of automation in raw material extraction and manufacturing became, the more the demand for labour in these sectors declined. Until recently the service sector has hardly been affected thanks to the protectionism by mankind's unique proposition in terms of phantasy, creativity, out of the box thinking, knowledge management, just to name a few.

1.1 Problem formulation

In the course of the last century the economy experienced a gradual shift from the primary and secondary to the tertiary sector. With the emergence of disruptive IT the service sector's traditional barriers are experiencing a fundamental erosion. In this context “disruptive IT” stands for big data, predictive analytics, machine learning and artificial intelligence. Robots, cyborgs and other kinds of machine replica of organic creatures are in scope as far as IT is involved to run their features either on a standalone or on a hybrid base.

The primary factors of production typically comprise land, labour and capital. These are transforming resources which in contrast to transformed resources are not part of the output. The demand for labour used to be a derivative of macro-economic cycles. In the phase of economic upswings, joblessness tended to vanish as a result of the economy's push toward full employment. On the contrary, a phase of economic downturn was typically accompanied by mounting unemployment rates.

However, there are strong signals that nowadays this traditional correlation has weakened[1]. The primary production factor of labour which this thesis refers to as human capital experiences a substitution by demand for that portion of capital that reflects the technological production factor of automation.

1.2 Objective of the Master Thesis

Therefore, the overall hypothesis this thesis seeks to verify is “disruptive IT will substitute human capital to an extent that will irreversibly materialize in an unprecedented technological rate of unemployment’. To operationalise this hypothesis questions such as the following shall be discussed in more detail:

- Which job profiles and industries could be affected the most and the least by disruptive IT?
- What criteria distinguish jobs in a certain geography, industry and business model from each other in terms of their substitutability by disruptive IT?
- What are disruptive IT's economic consequences with special emphasis on the labour market? Do individuals and the society overall need to reckon with the phenomenon of mass unemployment? How might contingency measures look like (e.g. reskilling and upskilling)?
- Which job profiles in existing as well as new economic sectors could emerge out of technological advances and mitigate joblessness? What are remaining entry barriers for disruptive IT in competition with human capital, if any?
- How may fiscal and monetary policies respond to the erosion of traditional cause- effect assumptions in assuring full employment? Which long-term strategies could be applied (e.g. basic education) and which short- to medium-term tactics could serve to seize the transformative opportunities and mitigate negative consequences?

1.3 Course of investigation

The answers to the aforementioned questions shall serve as valuable contribution to the ongoing debate about individual and societal effects of today's IT advances. The hypothesis shall be verified by means of research focusing on the analysis of academic literature and scientific debates as well as contemporary articles and empirical studies.

The investigation commences with an analysis of unemployment statistics and a trend analysis over the last decades with the objective to reveal what could be perceived as technological hike of the natural rate of unemployment. It will be shown that historical data yet does not provide sufficient evidence to become scientifically convinced about the hypothesis laid out before[2]. This does not come with surprise given the disruptive nature of recent IT advances which have yet only moderately started to unfold their full potential. Hence, disruption makes past-oriented statistics a second best guess of what to expect in future but suggests a rather simulation-based approach validated by expert opinion.

Therefore, we take a look at the contemporary study and analytical model of the Oxford researchers Carl B. Frey and Michael A. Osborne followed by the criticism of their criteria- based outside-in assessment on behalf of the Organisation for Economic Co-operation and Development (OECD). Next part of the discussion is the analysis of a recent report by the World Economic Forum (WEF) based on a self-assessment of industries' opinion leaders and field matter experts captured via means of employer survey. Finally, both the common denominator of as well as the key separators between the discussed material get consolidated and interpreted - with the distinction of the latter into those that arise out of methodological heterogeneity in contrast to those which relate to results and conclusions.

It's not academics' primary concern to make predictions and surely not part of their routine tasks. However, if academic research wants to make a contribution to the way policymakers, employers and employees as well as the society as a whole shape the future than it is reasonable to move out of the comfort zone and rethink the frontiers of academically serious work. Having said that, this master thesis has been dedicated to contribute in this spirit which seems to be supported by the courage of recent and ongoing technological advancements, e.g. predictive analytics that - thanks to big data - attempts to go beyond the status quo and draw a picture of the future based on captured patterns.

Hence, this master thesis wants to serve with quantified substance as facilitator in nudging science to expand the comfort zone.

2 Glossary of terms

The title of this thesis “The Disruptive Substitution of Human Capital by IT” mentions some termini technici that shall be briefly defined in order to lay the groundwork for the discussion of the subject matter going forward.

2.1 Substitution

The S curve concept is common knowledge in economic theory and best practice to visualize substitution (see Figure 1). As laid out before[3]this thesis assumes the disruptive shift in demand from the production factor of labour to capital, more precisely from human capital to IT, hence, giving rise to joblessness that is going to set and retain a level far away from the state of full employment.

illustration not visible in this excerpt

Figure 1: The S curve concept[4]

In this thesis's context let us expand the S curve concept from technologies as subsets of the production factor of capital to the primary factors of production themselves, i.e. land, labour and capital. Then the S curve on the left-hand side of Figure 1 may be deemed human capital. Incremental innovations are moves along the S curve whereas what this thesis is elaborating on are disruptive[5]innovations that create a new S curve such as the one to the right that we virtually want to label with “IT”. In a transitory period of substitution both curves may complement each other. Typically such a co-existence is temporary for the X coordinates the two S curves have in common and once having passed a “point of singularity” ends up with the previous factor's replacement through the new one. Such a development would be in the spirit of this master thesis's hypothesis, i.e. “disruptive IT will substitute human capital to an extent that will irreversibly materialize in an unprecedented technological rate ofunemploymenť.

The substitution of jobs and tasks is synonymously used for the substitution of human capital. When jobs or tasks are said to be substituted, what is actually meant is they do not require human capital anymore but get carried out by automatized means of IT.

In fact, the substitution of human capital does not come without the creation of new jobs in growing and even new industries. It is questionable how long newly created jobs are going to sustain and even compensate for the loss in jobs by IT substitution. As we will see the research consensus is rather bearish in regards of the compensatory potential of future job creation. What seems a given is a further tightening in the war for talents especially in these new occupations. The World Bank suggests the analogy to the razor- blade business model by quoting the spin given by Google's chief economist Hal Varian back in 2014: “Be an expensive complement (stats knowhow) to something that’s getting cheaper (data).”[6]

2.2 Disruption

In the connotation with the title of this thesis disruption is a quality attributed to substitution. It refers to a discipline of change that follows an exponentially shaped trajectory. Disruption inevitably prompts feelings of uncertainty, i.e. risks which have not yet been captured by metrics to make them measurable. It leaves a taste of feeling paralyzed to witness change with only mediocre control of what is going on. In a world of disruptive change a phrase in financial advisory used as standard disclaimer for clients' investment decisions and asset allocation ought to manage expectations: "The past is no predictor of the future!"

illustration not visible in this excerpt

As illustrated by Figure 2, beyond shorter time to innovation, disruptive change materializes in the acceleration of technological utilization. To penetrate a market of 50 million consumers it took the radio some 38 years, whereas TV required comparably low 13 years and the internet merely needed three years[7]. In essence, both technological innovation as well as utilization have become as short as multiple occurrences within the average demographic span of a human life. Notably, new technologies have evolved along the path of mobility and digitization toward the discipline of information technology.

2.3 Information Technology

Content-wise IT also known as Information and Communication Technology (ICT) can be related to five manufacturing and seven service sectors and sector aggregates respectively as specified in the United Nation's international standard industrial classification of all economic activities (ISIC Rev. 4)[8]:

illustration not visible in this excerpt[9]

Table 1: ISIC levels and nomenclature of IT-specific sectors[10]

The above listed items constitute the fundament for what this thesis grasps subject to IT as set forth previously[11], i.e. big data, predictive analysis, machine learning, artificial intelligence, robotics and robo advisory. As the studies discussed in this thesis use some of these terms literally, we will find exactly these terms being used synonymously for IT.

Process-wise IT may be deemed a trend and - as such - part of megatrends as they are predicted and promoted by various R&D facilities as well as consultancy companies under working titles such as “Connectivity”[12]or “Digitisation”[13]. Under “Connectivity” the German Zukunftsinstitut does for instance combine IT-related domains such as industry 4.0[14], big data, predictive analytics, internet of things and social networks.

2.4 Unemployment

Though unemployment does not find explicit mention in the title of this thesis its structural form reflects the nucleus of this work's hypothesis being discussed throughout all chapters henceforth. Systemic, sustainable and irreversible changes of the labour market are expected to materialize in structural unemployment, specifically in technological unemployment, as well as in underemployment. Therefore it appears appropriate to elaborate on the notion of structural joblessness, technological joblessness, underemployment and first and foremost the natural rate of unemployment.

The natural rate of unemployment is the rate of unemployment when the labor market is in equilibrium[15]. It is the difference between those who would like a job at the current wage level and those who are willing and able to take a job. The natural rate of unemployment will therefore include both frictional as well as structural unemployment and more specifically technological unemployment. The natural rate of unemployment is sometimes known as the “Non-Accelerating Inflation Rate of Unemployment” (NAIRU).

“Structural unemployment comes about in the long run... structural shifts affect the composition of labor-skill requirements. There is no problem as long as the laborforce is itselfable to adapt to the new requirements. ”[16]

Gilpatrick's definition emphasizes any time-pressure on workers' adaptability as structural unemployment's conditio sine qua non. This is particularly crucial once we believe to work in a world of disruptive change. In this case, the trade-off between speed of change and the human workforce's adaptability to this change becomes inevitable and typically materializes for job seekers who fail to get a job because they do not supply the skills in demand.

The honorable and renowned John Maynard Keynes once predicted technological unemployment as follows:

“due to our discovery of means of economising the use of labour outrunning the pace at which we can fnd new uses for labour”[17].

Again there is a timely dynamic embedded as defining element which puts pressure on workers' adaptability to change and insofar allows technological unemployment to be deemed a subcategory of structural unemployment with the speed of change accelerated specifically by technological advances in general and information technologies nowadays more than ever before.

The International Labour Organisation (ILO) has defined underemployment:

“when a person's employment is inadequate in relation to specified norms or alternative employment, account being taken of his or her occupational skill (training and working experience). Two principal forms of underemployment may be distinguished: visible and invisible.

... Visible underemployment is primarily a statistical concept directly measurable by labour force and other surveys, reflecting an insufficiency in the volume of employment.

... Invisible underemployment is primarily an analytical concept reflecting a misallocation oflabourresources ora fundamental imbalance as between labour and other factors ofproduction. Characteristic symptoms might be low income, underutilization of skill, low productivity. Analytical studies ofinvisible underemployment should be directed to the examination and analysis ofa wide variety ofdata, including income and skill levels (disguised underemployment) and productivity measures (potential underemployment).

... For operational reasons, the statistical measurement of underemployment may be limited to visible underemployment. ”[18]

For the purpose of this master thesis the definition of invisible underemployment is of pivotal relevance. This is firstly because invisible underemployment does account for substitutional effects of human capital by other factors of production such as capital invested into automation and digitization. Secondly, invisible underemployment is conditional to skills mismatches[19]. Underemployment may be interpreted as a hybrid form of full employment and structural unemployment or alternatively as an imperfect labor market solution to the woes of structural unemployment. It is this kind of grasp that lets the definition of underemployment belong to this very chapter titled “Unemployment”.

3 Main part

Starting point of the main part will be the description of the current state of global unemployment and a retrospective analysis with special consideration of the assumed dependency of job losses among human workforce due to digital disruption. For this, we will focus on historic and current numbers.

After that, the analysis will be expanded to the evaluation of future job losses as a result of IT substitution. All other things equal, population growth, let alone the number of jobs etc. form part of the ceteris paribus assumption, so demographic forecasts remain out of scope.

The future substitution of human capital by IT will be evaluated by three contemporary research studies, i.e. two predictive models and one employer survey. The two models are Frey's and Osborne's occupation-based simulation from 2013 and Arntz et al.'s task- based successor. The survey was conducted on behalf of the World Economic Forum in 2015 and allows a projection of opinion leaders' expectations until 2020.

3.1 Unemployment statistics and trend analysis

This chapter follows a three-step approach: First, the investigation focuses on global unemployment over the last few decades as well as on joblessness in the largest economies and economic unions in terms of their share in world population as a proxy for their portion in the working population around the globe[20], i.e. China, India, the EU, and the U.S. which together represent almost half of the world population. Second, research is carried out on global indices that measure IT advances in the broader sense of this thesis, such as a global index on digitisation, artificial intelligence, big data, etc. The purpose of step two is to clarify which metrics best serve the measurability of disruptive substitution of human capital. Alternatively, indices specific to the largest economies and economic unions form the scope of the investigation.

Thirdly, we contextualize indices on IT advances from step two and unemployment figures from step one and discuss them both graphically and via regression analysis. In the telos of this master thesis's hypothesis[21] what we would expect is an increasingly synchronised evolution of the two forces the more the substitution of human capital by IT unfolds its disruptive potential. Given this expectation the discussion of the correlation coefficient is supposed to serve its evaluation along the continuum from academic speculation to empirical proof.

illustration not visible in this excerpt

So let us turn to step one: the graph in Figure 3 represents the global unemployment rate according to the International Labour Organization's (ILO) standards. Data is publicly accessible for almost a quarter of a century, which provides a number of data points that suffice statistical representativeness. Looking back at the era of the First Gulf War in the early 1990s, the first record amounted to 6.3% unemployment among the global labour force[22].

As underscored by the trend line, the relative number of jobseekers around the world actually follows a downward trend with a record low in 2007 at a rate of 5.5%. Following the financial turmoil of the years 2008 and 2009, joblessness has found its way back to the pre-crisis trajectory.

illustration not visible in this excerpt

Figure 4: The unemployment rates 1991-2014 for the world’s largest economies and economic unions

Figure 4 provides a cascade of the global view down to the level of the largest economies and economic unions, which allows the following insights: Though unemployment in China is in sync with the global downward trend, most recent years repeat themselves in an adverse evolution, which can primarily be traced back to a slowdown of China's economic growth. India's unemployment keeps decreasing within a relatively narrow corridor of marginal 0.9%, i.e. from a minimum of 3.5% to a maximum of 4.4%.


[1]McKinsey (2014)

[2]see chapter 1.2 „Objective of the Master Thesis“

[3]see chapter 1.2 „Objective of the Master Thesis“

[4] Lüthje, Professional MBA business core module class 2, 25-26 February 2016, p. 10

[5]see the definition of disruption in chapter 2.2 “Disruption”

[6] Hal Varian quoted in The World Bank (2016), p. 130

[7] Dobbs, Manyika and Woetzel (2015) quoted on content/uploads/Disruption-2.jpg

[8] cf. European Commission (2016a)

[9] The ISIC level is determined by a cascade from alphanumerical sections to numerical divisions that are further divided into groups and classes. E.g. sector G/46/5/1 belongs to section G, division 46, group 5, class 1.

[10] cf. United Nations (2008)

[11]see chapter 1.1 “Problem formulation“

[12] cf. Zukunftsinstitut (2015)

[13] cf. Hansen (2016)

[14] The vogue expression “Industry 4.0” stands for fourth industrial revolution. The first one was witnessed in the late 18th century through the introduction of mechanical production plants. The early 20th century was the cradle of the second revolution when mass production and work specialisation found their way into the secondary economic sector. In the 1970s automation by IT triggered the third revolution. The fourth revolution is happening now characterized by the connection between physical and digital systems and stronger interdependencies between production, IT and the internet.

[15] cf. Friedman (1968), pp. 8

[16] Gilpatrick (1966), p. 204

[17] Keynes (1930), p. 3 quoted in Frey, Osborne (2013), p. 2

[18] ILO (1995), p. 32

[19]This will be elaborated on in more detail in chapter 3.4 “The WEF's “The Future of Jobs” Report”

[20]World Bank,

[21]see chapter 1.2 „1.2Objective of the Master Thesis“

[22]World Bank, UEM.TOTL.ZS.

Excerpt out of 59 pages


The Disruptive Substitution of Human Capital by IT
Catalog Number
ISBN (eBook)
ISBN (Book)
File size
1126 KB
Unemployment, Underemployment, Workforce, Labor, Capital, Substitution, Human Capital, IT, Frey, Osborne, Oxford, OECD, WEF, Employment, Future
Quote paper
Mag. Peter Weber (Author), 2017, The Disruptive Substitution of Human Capital by IT, Munich, GRIN Verlag,


  • No comments yet.
Read the ebook
Title: The Disruptive Substitution of Human Capital by IT

Upload papers

Your term paper / thesis:

- Publication as eBook and book
- High royalties for the sales
- Completely free - with ISBN
- It only takes five minutes
- Every paper finds readers

Publish now - it's free