TECHNOLOGY AT WORK
1Oclobi
1Oclobi
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
60<br />
Citi GPS: Global Perspectives & Solutions February 2015<br />
Industries at Risk<br />
Intelligent robots are now transforming<br />
industries such as construction, healthcare,<br />
professional services and finance<br />
The expanding scope of automation promises future gains in productivity in a wide<br />
range of industries. In the same way the mechanised assembly line transformed<br />
manufacturing, allowing companies to substantially cut production costs,<br />
sophisticated algorithms and robots with enhanced dexterity are now transforming<br />
industries such as construction, healthcare, professional services and finance.<br />
To estimate the industry employment impact the expanding scope of automation<br />
may have, we matched our occupation level data to the North American Industry<br />
Classification System (NAICS). 82 The susceptibility of various industries to<br />
automation thus reflects the intensity of different occupations and their probability of<br />
automation. For example, while all Tax Examiners work in the Government industry,<br />
only 44% of Computer Programmers work in Professional, Scientific and Technical<br />
Services. In fact, most occupations span across a wide range of industries.<br />
Figure 48. Employment share at risk by industry<br />
Low Risk (%) Medium Risk (%) High Risk (%)<br />
Accommodation & Food Services 2.8% 10.5% 86.7%<br />
Administrative & Support Services 1.6% 36.2% 62.2%<br />
Agriculture, Forestry, Fishing & Hunting 75.6% 12.0% 12.3%<br />
Arts, Entertainment & Recreation 47.9% 12.5% 39.6%<br />
Construction 21.6% 19.8% 58.6%<br />
Educational Services 63.1% 19.7% 17.2%<br />
Finance & Insurance 28.9% 17.3% 53.7%<br />
Government 46.2% 30.6% 23.2%<br />
Health Care & Social Assistance 39.4% 25.0% 35.6%<br />
Information 51.6% 38.3% 10.1%<br />
Management of Companies & Enterprises 82.8% 6.2% 11.0%<br />
Manufacturing 19.9% 18.4% 61.7%<br />
Mining, Quarrying and Oil & Gas Extraction 7.8% 46.3% 45.9%<br />
Other Services (ex Public Admin) 44.9% 24.7% 30.4%<br />
Professional, Scientific & Technical Services 54.0% 10.9% 35.1%<br />
Real Estate and Rental & Leasing 0.7% 32.0% 67.2%<br />
Retail Trade 14.5% 18.9% 66.6%<br />
Self-Employed 60.4% 8.9% 30.7%<br />
Transportation & Warehousing 5.5% 19.4% 75.0%<br />
Utilities 40.3% 27.8% 31.9%<br />
Wholesale Trade 15.9% 18.4% 65.7%<br />
Source: Oxford Martin School<br />
The susceptibility of computerisation varies<br />
substantially across industries<br />
As shown in Figure 48, the susceptibility to computerisation varies substantially<br />
across industries. In Accommodation & Food Services, as many as 87% of workers<br />
are at risk of automation, while only 10% of workers in Information are at risk.<br />
Although several occupations in Information, such as Motion Picture Projectionists<br />
and Broadcast Technicians are highly susceptible to computerisation, these only<br />
constitute a fraction of the industry’s total employment: 8,000 and 27,000 jobs,<br />
82 This analysis considered total employment of 98 million workers, against the 138<br />
million considered in Frey and Osborne (2013). Available data from O*NET records<br />
employment in an industry if at least 10% of employment in an occupation is engaged in<br />
that industry, leaving some employment uncounted. For example, the breakdown for<br />
Electrical Engineers records that 38% work in Manufacturing, 31% work in Professional,<br />
Scientific and Technical Services and 10% work in Utilities, such that 21% of<br />
employment for Electrical Engineers is distributed amongst the other industries in an<br />
unspecified way. We assume that this unspecified distribution is not reflective of<br />
computerisability and that it does not introduce any systemic risk.<br />
© 2015 Citigroup