One might question if the seeming decoupling of labor and growth, the recent rise in inequality, the reduction of labors share of income cannot be explained by another global macro trend, namely the offshoring of low paying factory jobs. Surely this trend must have had an effect on for example inequality in the US, as the American factory worker now has to compete with the Chinese. But in fact labors share of income has been falling globally, even in China to where these low paying jobs have moved. This can be seen in figure 19. If offshoring was the full explanation, you would not expect the phenomenon of labor and growth decoupling to be global in scope.
In conclusion of this chapter on the evidence for the decoupling of labor and income, this thesis has shown that certain technologies are increasing exponentially and that labor growth per capita has stalled while GDP per capita has continued to rise. This fits with the statistics that show how society has become more unequal over time. Also it has been shown how 𝛼, the elasticity of capital that can be seen as a proxy for the pace of
technological change has been rising steadily over the past three decades, and finally how the return on labor is as a percent of total income has been falling consistently in countries around the world. Hopefully all these pieces of evidence together create a compelling argument for believing that a fundamental shift in technology’s impact on labor is occurring.
14 Changes in demanded work skills
Now that the case for why something fundamental could be changing the job market has been made, this thesis will examine what effects this has had on the skills required in jobs.
I thus searched for job market data in databases in the developed world. My first choice for a country was the US, because it is large enough to alleviate any doubts that the
56
findings here might not represent the bigger picture (as opposed to selecting German or UK data). The hundreds of lines of code and database work done accounted for about half of the numerous hours that went into this thesis. The computationally most expensive statistical calculation involved more than 20 million data points and took more than 24 hours to complete (as is also discussed in detail in the “statistic methodology” part of this paper).
As a quick recap of the statistical methodology chapter, data from a public periodic job market survey called “O*NET” was used. More specifically it was examined statistically what skills each job position required, and how the required skills had evolved over time.
Additionally, the majority of skill trends are statistically significant. The table on the next page displays these results.
Before the analysis was carried out, I expected the innovative, agile and tech type skills to increase in demand. The “Silicon valley” type jobs if you like. On the other hand, I had expected certain repetitive skills to be more and more automated and therefore less in demand.
As can be seen from the actual results in table 3, reality is not that simple. Table 3 measures not only how much the demand for certain skills have changed year over year, but also how much pay has changed for each skill.
57
Table 3 – The trend of change of demand and pay respectively, for different skills used on the job market. The table is sorted and colored according to demand change.
Skill Description Relative Pay
Change
Demand Change
Negotiation Bringing others together and trying to reconcile differences. 1.59% 5.43%
Persuasion Persuading others to change their minds or behavior. 0.07% 4.73%
Management of Personnel Resources
Motivating, developing, and directing people as they work, identifying the best people for the job.
2.35% 4.59%
Service Orientation Actively looking for ways to help people. -0.12% 3.41%
Social Perceptiveness
Being aware of others' reactions and understanding why they react as they do.
0.35% 3.05%
Operation Monitoring
Watching gauges, dials, or other indicators to make sure a machine is working properly.
-0.54% 2.98%
Time Management Managing one's own time and the time of others. 1.07% 2.72%
Instructing Teaching others how to do something. 1.12% 1.58%
Monitoring Monitoring/Assessing performance of yourself, other individuals, or organizations to make improvements or take corrective action.
0.32% 1.47%
Critical Thinking Using logic and reasoning to identify the strengths and weaknesses of alternative solutions, conclusions or approaches to problems.
1.25% 1.35%
* Programming Writing computer programs for various purposes. 3.57% 1.27%
Coordination Adjusting actions in relation to others' actions. 1.16% 0.90%
Active Listening Giving full attention to what other people are saying, taking time to understand the points being made, asking questions as
appropriate, and not interrupting at inappropriate times.
0.78% 0.76%
* Judgment and Decision Making
Considering the relative costs and benefits of potential actions to choose the most appropriate one.
0.95% 0.65%
* Speaking Talking to others to convey information effectively. 0.89% 0.52%
* Learning Strategies Selecting and using training/instructional methods and procedures appropriate for the situation when learning or teaching new things.
0.74% -0.09%
* Active Learning Understanding the implications of new information for both current and future problem-solving and decision-making.
1.45% -0.22%
* Writing Communicating effectively in writing as appropriate for the needs of the audience.
1.42% -0.45%
* Reading
Comprehension
Understanding written sentences and paragraphs in work related documents.
0.97% -0.88%
58
* Troubleshooting Determining causes of operating errors and deciding what to do about it.
-0.06% -1.08%
Management of Financial Resources
Determining how money will be spent to get the work done, and accounting for these expenditures.
2.74% -2.02%
Operations Analysis
Analyzing needs and product requirements to create a design. 1.17% -2.16%
Repairing Repairing machines or systems using the needed tools. -0.67% -2.66%
Operation and Control
Controlling operations of equipment or systems. -0.91% -2.78%
Equipment Maintenance
Performing routine maintenance on equipment and determining when and what kind of maintenance is needed.
-0.95% -2.99%
Management of Material Resources
Obtaining and seeing to the appropriate use of equipment, facilities, and materials needed to do certain work.
1.55% -3.08%
Science Using scientific rules and methods to solve problems. 1.35% -4.00%
Mathematics Using mathematics to solve problems. 0.51% -4.18%
Technology Design Generating or adapting equipment and technology to serve user needs.
0.56% -6.29%
Equipment Selection
Determining the kind of tools and equipment needed to do a job. -0.65% -8.27%
Installation Installing equipment, machines, wiring, or programs to meet specifications.
-1.67% -9.03%
From looking at the data it becomes clear that interpersonal skills are increasingly in demand. Negotiation, persuasion, management of personal resources, service orientation and social perception all top the list of skills trending upwards. On the other hand, a lot of the technical skills are less and less required, including science and mathematics. This trend seems somewhat counter to the common narrative about how automation will affect the job market. But if automation is increasing so fast that it is even outcompeting labor in the industries that is created around it, this trend is actually what you would expect to see.
It might be that technology and devices are becoming simpler and easier to use for the end user, more “plug and play”, whereas it used to require more professional expertise to set up. We have seen some of this with consumer products: Whereas users 10 years ago were
Demand change: > 6% (unused) 6% to 3% 3% to 0% 0% to -3% -3% to -6% -6% >
* means statistically insignificant (P > 0.05)
59
expected to occasionally find and install drivers on their own, using modern tablets or smartphones, user are hardly expected to do more than “one click” install application from an “app store”.
Although an effort has been made to make the data in table 3 above as easily digestible as possible, it is still too much information to make full sense of at a glance. In order to condense the information enough to draw a picture of general trends, the skills have been divided into three categories: “Interpersonal skills”, “agile, innovative and tech skills” and
“operations and maintenance skills”. The reasoning behind choosing these three specific categories is based on the specific skills measured in the O*Net data, but also on the way they reflect the societal macro trends that after much scrutiny is the my main take away from the full data points. Of course, by introducing these categories a subjective factor is also being introduced into the analysis of the data, and the summery in table 4 below should therefore be taken as such. In table 3 it is made clear what skills have been
assigned to which category by coloring the space before the skill name. The colors reflect categories as such:
Operations and maintenance skills Interpersonal skills Agile, innovative and tech skills
The more comprehensive data points that this summary is derived from, as well as decisions taken about how to categorize each skill, can then be scrutinized using this coloring data in table 3.
Table 4 - Skill groups Average Demand Change
Income index Pay Change Index
Interpersonal skills 2,57% 100 100
Agile, innovative and tech skills -1,18% 117 138
Operations and maintenance skills -2,75% 80 -5
What this summarized form of the data shows more clearly is that American workers are more than anything gravitating towards positions with a higher level of interpersonal skills
60
and away from other categories. At the same time skills like “agile, innovative and tech skills”, the set of skills that are probably most commonly regarded as the jobs of the future, are decreasing in demand, but increasing more than the other categories in pay.
The income index shows how income is related to each of the three categories in the 2013 data, and the category Interpersonal skills is here set to 100. As can be seen from this, even though “Agile, innovative and tech skills” are the best compensated and fastest rising category in terms of pay, it is shrinking in terms of demand. This seems to support the notion that the jobs and skills required by those who build automation are themselves being automated faster than they are emerging.