Retail Apocalypse Postponed Not Cancelled

• Brian Schaitkin
• Brian Schaitkin

The share of e-commerce in total retail sales has grown from 2.3 percent to 9.1 percent since 2005 during a period where “new” retail sectors have created just 749,000 jobs.

Note: New portion of “old and new” retail consists of couriers and messengers, warehousing and storage, and electronic shopping and mail-order houses.

Sources: Census Bureau, BLS, calculations by The Conference Board

During this same period, 749,000 new jobs were created in the out-of-store portion of “old and new” retail.  If one assumes that every 6.8 percent increase in e-commerce yields 749,000 new jobs, a 100 percent online retail sector would employ just 12.1 million workers compared to 17.6 million today.  This assumes that productivity in “old and new” retail grows no faster than the growth of consumer demand for goods.

How quickly will e-commerce grow as a share of retail sales and how soon will employees in “old and new” retail be displaced? Under an “Apocalypse is an Exaggeration” scenario, e-commerce will grow at a modest pace as a share of total retail sales because many of the lowest hanging e-commerce fruit have already been plucked. Therefore, the transformation from in-store retail to e-commerce would take quite a long time. The logistical challenge of shipping books to individual customers is orders of magnitude less complicated than delivering groceries. The “in-store” experience also provides special value to some consumers, especially when the ability to touch and feel merchandise and consult with experts is part of the value proposition stores and their workers provide. Firms will learn how to adapt to the challenge of e-commerce in less easily adapted industries, but because of these challenges, growth in the e-commerce share may increase by only 6.8 percent every 13 years, as happened between 2005 and 2017.  Under this scenario, the share of e-commerce in retail sales would be 21 percent by the end of 2040, and the “old and new” retail sector will employ 16.8 million workers, compared to 17.6 million today.  A disappointing job trajectory to be sure, but hardly apocalyptic.

Apocalypse is an Exaggeration: If e-commerce grows slowly as a share of total retail sales, the number of jobs in “old and new” retail will decline modestly over the long-term.

Sources: BLS, Census Bureau

The 2005 to 2017 period though lights the way to a bolder, yet in my view more likely “Apocalypse is Ongoing” scenario. This scenario assumes that exponential growth of the e-commerce share will continue as it did from 2005 to 2017 meaning that the e-commerce share of retail will double every 6.5 years. Incentives for firms to adapt new sectors to e-commerce will be tremendous due to the convenience and efficiency commerce without stores can provide. Companies can apply the lessons from developing one form of e-commerce to new product areas with knowledge of the pitfalls they are likely to encounter. Under this scenario, e-commerce would represent 18 percent of retail sales by the middle of 2024, 36 percent by the end of 2030, and 100 percent by the middle of 2040.  Initially, job losses resulting from the shift to e-commerce would be small in “old and new” retail with 738,000 jobs disappearing between now and the end of 2025.  By 2040, however, only 12.1 million workers, all employed by “new” retail sectors, would be able to manage all retail sales activities, a loss of 5.5 million jobs. 

“Apocalypse is Ongoing”: As the share of e-commerce expands more quickly, the number of jobs in the “old and new” retail sector will decline rapidly.

Note: “Old” retail employment includes both in-store retail and other non-store retail

Sources: BLS, Census Bureau

An even more dramatic change is quite possible where even fewer jobs will be created in “old and new” retail as e-commerce expands its reach. Under an “Apocalypse has Arrived” scenario, the potential for autonomous vehicles to streamline deliveries, and additional advancements in robotics to help ultra-modern warehouses move goods in and out even more efficiently, would drive further productivity improvements. The next recession would likely catalyze shifts already ongoing in both in-store retailing and e-commerce, driving employment in “old and new” retail down rapidly.

The “new” retail sector includes many of the economy’s most innovative firms who are likely to develop and deploy technology that could result generate these rapid productivity gains.  If the experience of late 1990s manufacturing sector is any guide, expect “old and new” retail job losses to exceed those projected in the above “Apocalypse is Ongoing” scenario.

Finally, network effects may provide a further source of acceleration for e-commerce growth. As e-commerce in a given sector increases as a share of total sales, surviving in-store retail branches could become progressively harder to maintain.  Downstream supplier networks could wither and may also become increasingly well adapted to e-commerce. Synergies between e-commerce activities in different industries could also encourage faster growth of the e-commerce share generally. The quicker e-commerce grows, the more rapidly the number of jobs in “old and new” retail would dwindle.

This post remains agnostic as to questions around total future employment and what exactly workers across the labor market will be doing. Labor market data though very clearly suggest that jobs in warehousing, delivery, and e-commerce, are unlikely to fully replace those lost in brick and mortar retail.