Christopher Andrews’ May blogpost takes up a question that is sure to get increasing attention: is automation and artificial intelligence going to significantly reduce the number of jobs? What is the future of work? Andrews looks at self-checkout lanes in supermarkets—one of the biggest and most obvious candidates for job elimination—and has surprising and very interesting findings. I won’t spoil things by revealing his answer. All I’ll say is that this is a must-read.
And on a personal note: Having written books called The Overworked American and The Overspent American myself, I’m delighted that Christopher’s forthcoming book is tentatively titled The Overworked Consumer.
– Juliet Schor
The End of Work (or More Work)?
By Christopher Andrews
Unless you grow your own produce, you’ve probably recently visited the grocery store, many of which increasingly offer self-checkout lanes where you can “do it yourself”. Similar forms of self-service technology can be found throughout the service industry (e.g., ATMs, ticketing kiosks), as companies increasingly beckon consumers to engage in what has been characterized by sociologists as “shadow work” or “prosumption”. Accordingly, critics predict that these technologies threaten to displace paid labor with the unpaid labor of consumers (or prosumers), resulting in significant job loss.
To a certain degree, these concerns reflect a recurring fear of technological unemployment; examples include “the great employment controversy” of the 1950s and “deskilling debate” of the 1970s. Yet, if technology eliminates jobs, economic historians note, it also creates opportunities and demands for others; for example, the decline of agricultural jobs coincided with the growth of the service industry, while similar declines in manufacturing jobs coincided with the digital revolution.
In some respects, these recurring fears of technological unemployment might be characterized as moral panics in which technological innovations are depicted by the mass media as folk devils that threaten the economic security of American workers. Recent bestselling books such as The Second Machine Age (2016), Rise of the Robots (2016), and Race Against the Machine (2012) depict not-so-distant futures in which automation results in job loss and technological unemployment. But if such technology threatens to eliminate work, as many suggest, why are there still so many jobs?
I decided to explore this question by examining the introduction of self-checkout lanes in the retail food industry. Why supermarkets? First, they employ a majority of the kinds of workers most likely to be affected by these kinds of self-service technologies. Supermarkets employ more cashiers than any other industry – almost twice as many as the next two industries combined – with cashiers representing nearly a third of all store employees. There are also strong incentives for supermarkets to automate by reducing labor costs. Unlike retail stores and fast food restaurants, supermarkets frequently employ unionized workforces, exposing them to rising labor costs and work stoppages; the 2003 UFCW strike cost the retail food industry an estimated two billion dollars. And with nonunion competitors (e.g., WalMart) entering the retail food industry, supermarkets face increasing pressures to reduce labor costs.
Yet, while the number of stores offering self-checkout lanes has increased dramatically – from six percent in 1999 to ninety-five in 2007, with most stores averaging four self-checkout lanes – there are actually more cashiers today than a decade ago, even when we control for the number of stores. To paraphrase the Nobel-prize winning economist Robert Solow, you can see the effects of self-checkout lanes everywhere except in the statistics.
One of the reasons why self-checkout lanes have not eliminated cashiers is theft. Despite built-in security measures, retailers reportedly lost 44 billion dollars in 2014 due to theft, with rates of theft reported to be as much as five times higher in self-checkout lanes compared to cashier-operated lanes, while a national survey revealed that one in five shoppers admitted to regularly stealing from self-checkout lanes.
Another reason self-checkout lanes are not eliminating jobs is because they cannot operate independently, a problem that appears to have plagued efforts to automate supermarkets from the very beginning. In 1937, Clarence Saunders attempted to further cut costs in his pioneering self-service stores by replacing cashiers with an automated checkout system, yet customer difficulty, mechanical breakdowns, and soaring labor costs for electricians led him to close the store after just one year. Today, technical problems continue to limit stores’ ability to fully automate; self-checkout lanes unexpectedly “freeze”, unwanted items have to be manually voided, and paper for receipts has to be replenished.
Yet, perhaps the most important reason self-checkout lanes have not eliminated jobs is because of customer service. Home Depot learned this lesson the hard way when former CEO Robert Nardelli attempted to cut costs by replacing cashiers with self-checkout lanes; frustrated customers fled to archrival Lowe’s whose stock price subsequently doubled while Home Depot slipped to last among major U.S. retailers in the University of Michigan’s annual American Consumer Satisfaction Index. Profit margins in the retail food industry are also razor-thin – often less than one percent due to the abundance of competition – so stores carefully limit and monitor their use for fear of losing customers.
In short, whether self-checkout lanes and similar forms of self-service will eliminate jobs depends on a series of technical, social, and political factors. Many supermarkets currently employ unionized workforces that have collective bargaining agreements stipulating that employees may not be replaced by machines; however, growing public acceptance of self-service technology – as well as increasing competition from non-union competitors like WalMart – may erode unions’ ability to secure such provisions. Moreover, surveys of younger shoppers find an increasing willingness and desire for self-service, suggesting that younger cohorts may be more willing to “do it yourself” precisely because they have been socialized to do so from an early age.
While the introduction of self-checkout lanes may not be eliminating jobs, they have contributed to an expansion of unpaid work. Similar to the “second shift” that working women face at home, the growing use of self-service technology may help to explain why Americans increasingly report feeling overworked despite empirical evidence to the contrary, as unpaid work encroaches on what little remaining leisure time Americans enjoy.
Indeed, it would seem that the notion that “the customer is always right” is quickly being supplanted by a notion that customers can (and perhaps should) “do-it-yourself”, an idea that fits well alongside American virtues of rugged individualism and self-reliance. Historically, for many Americans service has meant having someone else perform a job for you; with self-service, “service” may come to mean something Americans choose to do for themselves. Caveat emptor.
Christopher Andrews is an Assistant Professor of Sociology at Drew University. He is currently working on a book about self-service, consumers, and unpaid work tentatively titled The Overworked Consumer under contract with Lexington Books.