Piggybacking on the research of academics and researchers who studied the impact of automation on jobs a few years ago, two insurance analysts recently calculated an important impact for workers compensation carriers—the possible disappearance of more than 40 percent of premiums by 2030.
In a September research report, William Wilt and Alan Zimmermann of Assured Research are quick to dismiss the dire forecast, suggesting that government intervention would stem the tide of significant job losses and that displaced workers would find new roles.
Nevertheless, taking a close look at the percentage of jobs characterized as “high risk” for nine specific industries included in a February 2015 report by Citi GPS and the Oxford Martin School at the University of Oxford, the insurance analysts apply the percentages to the number of employees in each industry, calculating that some 60.4 million jobs could be lost in these industries. Those lost jobs represent $26.5 billion of workers comp premium lost, for 42 percent of the nationwide premium total, the Assured Research analysts found.
The Citi/Oxford report titled, Technology at Work: The Future of Innovation and Employment, actually details the percentage of jobs that have “low risk,” “medium risk,” and “high risk” of automation for 21 different NAICS (North American Industry Classification System) codes (on page 60 of the report). The most endangered, according to Oxford Martin’s Carl Benedikt Frey and Michael Osborne, is the Accommodation and Food Services class, where 87 percent of workers are at “high risk” of having their jobs automated. In contrast, only 10 percent of workers in the “Information” NAICS code face a similar level of risk, they note.
The nine that Wilt and Zimmermann focused on were: Construction; Manufacturing; Administrative and Support Services; Transportation and Warehousing; Wholesale Trade; Retail Trade; Accommodation and Food Services; Finance and Insurance; Real Estate and Rental and Leasing.
The insurance analysts picked those particular classes after being intrigued by a July Wall Street Journal article that set forth the “high risk” percentages for those nine classes (“Robots Are Replacing Workers Where You Shop,” WSJ, July 19,2017). The focus of that WSJ article was the retail sector, where Osborne and Frey put 66.6 percent of jobs in the “high risk” category and Wilt and Zimmermann calculate a loss of $2.3 billion in workers comp premium associated with 11.4 million endangered jobs.
Of the nine classes included in the Assured Research workers comp analysis, the Construction class and Finance and Insurance class had the lowest percentages of jobs at risk—59 percent for Construction and 54 percent for Finance and Insurance. But while the Construction sector accounts for only 5.1 million of 60.4 million at-risk jobs across all nine classes analyzed, the sector accounts for the greatest loss of workers comp premium—$6.7 billion.
The remaining lost workers comp premium figures are available from Assured Research in the firm’s September 2017 report. The September edition also looks at the personal auto line and the potential impact of retrofitting cars for safety features on personal auto costs. In addition, the analysts include a historical look at mergers and acquisition in the Bermuda insurance market and in the reinsurance sector.
As for the workers comp analysis, although Wilt and Zimmermann stress that they believe many job-loss forecasts are too apocalyptic, they deliver an important takeaway message for insurers.
“The demand for property/casualty insurance should NOT be taken for granted. Complacent insurers not investing in the research to understand how economic, legal, technological, and social changes can affect their business will, over time, become extinct,” they write.
For insurance executives interested in the more direct impact of automation on positions in the insurance industry, an earlier report by Osborne and Frey published in 2013—”The Future of Employment: How Susceptible Are Jobs to Computerization?“—provides some estimates of automation risk by job title. That report, for example, listed insurance underwriting as the fifth most susceptible occupation to automation—coming in at No. 698 of 702 individual occupations studied, where high numbers indicate vulnerability. Insurance agents fared better in that analysis—ranked at No. 565.