The_Future_of_Employment
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
TABLE I. O∗NET variables that serve as indicators <strong>of</strong> bottlenecks to computerisation.<br />
Computerisation<br />
bottleneck<br />
Perception<br />
and<br />
Manipulation<br />
Creative<br />
Intelligence<br />
O∗NET Variable<br />
Finger<br />
Dexterity<br />
Manual<br />
Dexterity<br />
Cramped Work Space,<br />
Awkward Positions<br />
Originality<br />
Fine Arts<br />
O∗NET Description<br />
<strong>The</strong> ability to make precisely coordinated movements <strong>of</strong><br />
the fingers <strong>of</strong> one or both hands to grasp, manipulate, or<br />
assemble very small objects.<br />
<strong>The</strong> ability to quickly move your hand, your hand together<br />
with your arm, or your two hands to grasp, manipulate, or<br />
assemble objects.<br />
How <strong>of</strong>ten does this job require working in cramped work<br />
spaces that requires getting into awkward positions?<br />
<strong>The</strong> ability to come up with unusual or clever ideas about<br />
a given topic or situation, or to develop creative ways to<br />
solve a problem.<br />
Knowledge <strong>of</strong> theory and techniques required to compose,<br />
produce, and perform works <strong>of</strong> music, dance, visual arts,<br />
drama, and sculpture.<br />
Social<br />
Intelligence<br />
Social<br />
Perceptiveness<br />
Being aware <strong>of</strong> others’ reactions and understanding why<br />
they react as they do.<br />
Negotiation Bringing others together and trying to reconcile<br />
differences.<br />
Persuasion<br />
Assisting and Caring for<br />
Others<br />
Persuading others to change their minds or behavior.<br />
Providing personal assistance, medical attention, emotional<br />
support, or other personal care to others such as<br />
coworkers, customers, or patients.<br />
the level <strong>of</strong> “Manual Dexterity” computer-controlled equipment would require<br />
to perform a specific occupation. An exception is the “Cramped work space”<br />
variable, which measures the frequency <strong>of</strong> unstructured work.<br />
Hence, in short, by hand-labelling occupations, we work around the issue<br />
that O∗NET data was not gathered to specifically measure the automatability <strong>of</strong><br />
jobs in a similar manner to Blinder (2009). In addition, we mitigate some <strong>of</strong> the<br />
subjective biases held by the researchers by using objective O∗NET variables to<br />
correct potential hand-labelling errors. <strong>The</strong> fact that we label only 70 <strong>of</strong> the full<br />
702 occupations, selecting those occupations whose computerisation label we<br />
are highly confident about, further reduces the risk <strong>of</strong> subjective bias affecting<br />
our analysis. To develop an algorithm appropriate for this task, we turn to<br />
probabilistic classification.<br />
31