The reality of automation and its impact on the labour market is a lot less dire for workers than it seems

We have all seen the slew of news stories, headlines and internet memes saying a bump in the minimum wage will result in mass unemployment due to automation. Being replaced by robots it appears is the inevitable outcome of wage increases for workers. Since the Ontario Liberal government announced it would raise the minimum wage to $15/h by 2019, a number of high profile companies like Loblaw, Metro and Dollarama have announced they plan to increase the use of automation in response. Rather than pay increased wages to workers they will invest their profits in labour-saving technology, such as self-checkout machines.

But what, if any, truth is there to this? Are all workers destined to be replaced by machines? Does raising the minimum wage simply result in increased job loss due to automation?

On the surface this all looks quite grim for workers. But beyond the technophile narratives and splashy headlines the reality of automation and its impact on the labour market is a lot more complicated and a lot less dire for workers than it seems.

Agriculture to industry

The anxiety over automation and the transformation of work is nothing particularly new. The history of capitalism is marked by a constant revolution in the nature and organization of work. A hundred and fifty years ago the Canadian economy was overwhelmingly agrarian.

There were no cars, tractors, telephones, or even electrical grids in the country.

In 1881 48 percent of the total labour force was in the agricultural sector. In 1911 that number dropped to 34 percent and by 1971 it was at 5.6 percent. Today the percentage of the workforce in the agricultural sector is roughly 1.6 percent. The concentration and consolidation of powerful corporations in the agricultural sector, and the implementation of labour-saving technology especially after World War Two, led to the slow and steady decline of the agricultural workforce.

The reduction in the agricultural workforce due to automation did not lead to mass unemployment nor the end of work.

The decline of employment in the agricultural sector did however correspond with the rise of the industrial production. Increased capital and money was invested in larger and larger factories to produce mass goods. Early craft production of farm equipment, tools and consumer goods was already being replaced by methods of industrial production in the late 1800s.

With the rise of large-scale industrial production came the reorganization of work life. The norms and rhythms of agricultural and craft work with their high degree of independence and control over production were reorganized by capitalists and the newly forming managerial class required to manage sprawling new corporate empires.

By increasing the division of labour and introducing new technologies, management was over time able to deskill craft workers. By smashing the power of craft workers over production, management could cheapen the cost of labour and intensify production more easily. The introduction of new technologies was not simply about replacing human labour, although it could and did do that; it was reorganizing the pace and rhythms of production to increase output. This process was not determined by the existence of technology itself, rather it stemmed from the needs of capitalists to increase profits in competition with other capitalists.

Brakes on automation

Thus automation was and still remains chiefly a form of labour discipline to increase output and maximize profits. Firms that choose not to automate in order to ramp up output, are likely to lose market share because competitors can undercut them. Firms which can utilize new production methods and implement labour-saving technologies before other firms follow suit will have an added benefit of a competitive advantage in the marketplace.

However, while these market forces push firms to towards automation, there also market forces which act as brake on this process. When firms increase the amount of new machinery relative to labour costs (a process called “capital deepening”), this tends to drive down labour costs. But lower labour costs then slows the rate at which firms will aim to replace workers by machines.

Another brake on the pace of automation is that the fixed capital of machines is very inflexible. Firms pay for the costs upfront and risk the possibility that the technology will be rendered outdated or even obsolete by other technological advances. And the risks become massive if a downturn occurs. Machines cannot be fired and their cost remains on the books, depreciating in value even if they aren’t being used.

Finally, when the economy is in a low-growth period, as it is now, the tendency of firms to invest heavily into new technology to increase output is greatly reduced. There are entire sectors of the economy – like steel, rubber, textiles – where capacity far outsrips current demand.

Automation and the end of work?

It is true that over the last 150 years employers have reorganized and automated significant sections of production. The manufacturing sector has been dramatically transformed by the reorganization of the labour process and the prevalent use of labour-saving technology.

But it is also true that the predictions of a worklerless future, where automation has created mass unemployment, has likewise proven overhyped. Prominent labour activists, academics, political pundits and of course corporations have long been predicting the demise of work due to automation.

These predictions of the end of work rested on a profound misreading of how the economy operates. For a case in point about why technological deterministic arguments are wanting, one needs to look no further than the ATM bank machine.

The ATM automated simple functions for bank tellers. But today, statistics show in places like the United States, decades after the ATM was introduced, there are more bank tellers than ever. The ATM did reduce the number of tellers per branch from 21 to roughly 12, but that in turn reduced the cost of opening bank branches and as such branch expansion exploded (aided by other factors such as deregulation).

The advent of online banking, is probably a more significant factor in potential job shifts in the banking sector than the ATM, as online banking reduces the need for brick and mortar store fronts. But this also creates others jobs in web development, call centres, cyber security etc. In many ways online banking is a holdover of an industry slowly adapting to the 1995-2005 wave of technological restructuring.

Likewise the advent of computer software for bookkeeping saw a drastic fall in the amount of bookkeepers in the United States. But it also created the conditions for the number of accountants, financial managers and management analysts to grow precipitously .

The idea that automation equals job loss is not such a simple equation. Numerous other factors such as broader economic conditions, existing productive capacity, the unleashing of business expansion, shifting job tasks, the price of labour, and the role of state regulation all complicate the thesis of that automation means the destruction of jobs. When it comes to more complex service sector tasks like food service there is even more reason to be sceptical about how automatic kiosks and robots will create massive job loss.

$15 and fears of automation

There has been a move in fast food, retail and other related services to automate with machines. However, contrary to the hype, the results have not been so clear cut when it comes to job replacement.

In grocery stores automation in the form of self-service checkout machines have existed for 15 years and they have not seen widespread implementation nor major job loss. But new fears and hype are returning because of the push for a $15 minimum wage.

Metro is threatening more automation, but there seems to be little indication that grocery stores have figured out how to employ automation at the store level to significantly decrease the need for workers. The efficiencies of logistics tech for just-in-time delivery were already reorganized years ago. Sobeys has even acknowledged the reality that it can’t cut staff as a response to the minimum wage increase, because it is in the service industry they have factor in the quality of service they provide.

Fast food companies like McDonald’s, which does 70 percent of its business via drive-thru, have installed expensive kiosk machines but this has not resulted in massive job loss. It has in some places shifted some labour from front-end cashier work to the back-end kitchen, or even created the demand for table service.

In a recent CBC article about the rise of automation due to Ontario’s $15 minimum wage, a McDonald’s franchisee noted:

Some McDonald’s restaurants have introduced self-serve kiosks as part a broad ranging strategy to create a “restaurant of the future,” said Jason Trussel, who owns five McDonalds restaurants. He told CBC the machines are not going to replace human workers. In the past 18 months, his staffing levels have increased by 66 employees. Trussel introduced the kiosks about a year ago.

The machines simply give people more options, which brings in more customers and requires more staff, he said.

Some food chains have really tried to bring in new technology to reduce costs and increase output. Starbucks thought their mobile ordering app would rationalize sales and service. Instead, it has increased the need for staff, not lessened it. As one news report noted:

That’s because when customers order on their smartphones, during busy times of day, these orders are all immediately dumped on baristas. Then, when the hoard of customers show up to pick up their drinks at the exact same time, it creates bottlenecks and scares walk-in customers from even attempting to order.

The Starbucks mobile app has actually led to the hiring of more staff and the plans to open up more stores:

To address the problem, Starbucks has focused on workflow, including adding one or two more baristas dedicated to mobile orders at high-volume stores during peak hours. The company is also testing sending text alerts to customers when orders are ready.

Some on the RetailWire BrainTrust panel of retailing experts saw the mobile customer-only concept as a win with implications beyond Starbucks.

“Having one or more dedicated mobile order and pay-only stores, near existing dine-in Starbucks locations, will allow for the load-balancing of orders — smoothing out the experience for customers and baristas,” said Shawn Harris, North American retail and hospitality industry lead at Zebra Technologies. “Smart idea.”

Beyond the Hype Machine

Automation and the transformation of work is part of capitalism. It has happened, is happening, and will continue to happen. If we understand this, we should also ask ourselves another question, at what rate is it happening?

By all discernable macroeconomic measures automation is occurring at the slowest rate since World War Two. The rate of productivity growth since 2005, which measures output per worker, the rate of capital deepening, which measures the amount of capital per worker in a firm, and “job churn” measuring the rate at which workers move professions, are all showing statistically that automation is not accelerating, but actually crawling.

For workers the question becomes who does the automation narrative serve? At almost every instance it has been used as cudgel to beat back workers demands for better wages and working conditions. As workers we should recognize that there are possible transformative technologies on the horizon that could re-shape the world of work. Automation and the advent of new technology are not inherently good or bad, but when automation is driven by profit and competition and not human need and environmental sustainability, workers and the planet will pay the price.

Workers will see none of the benefits of automation unless we are able to exert our collective power. This means rejecting the false fears of the boss’s automation and demanding more control and power in the workplace.

Tagged under: