Academics and economists have repeatedly underestimated the impact that immigration and automation would have on the labor market. As data on productivity gains and labor-force participation clearly show, the notion that innovation ultimately creates jobs by allowing workers to focus on higher-level problems is an illusion. If it were true, then why aren’t we already seeing more of the 20 million prime-age men who have inexplicably dropped out of the labor force welcomed back in?
As we've noted time and time again, after decimating American manufacturing jobs in the 1990s, automation is now coming for service-industry workers like those in the retail and food-service industries. Earlier this week, we shared an analysis from Cowen that showed new kiosks being adopted by McDonald’s will result in the destruction of 2,500 jobs at its US eateries. And now, Bloomberg has published a “quick take” questioning this “official” narrative and pointing out the very real carnage that service sector workers are already facing. In it, the reporters noted how economists have repeatedly misjudged how our capacity to innovate would impact the labor market. For example, 13 years ago, two leading economists published a paper arguing that artificial intelligence would never allow a driverless car to safely execute a left turn because there are too many variables at work. Six years after that, Google proved it could make cars fully autonomous, threatening the livelihood of millions of taxi and truck drivers. And now Google, Uber, Tesla and the big car manufacturers are all exploring and testing this technology. Ford has said it plans to introduce a fully autonomous car by 2021.
“Throughout much of the developed world, gainful employment is seen as almost a fundamental right. But what if, in the not-too-distant future, there won’t be enough jobs to go around? That’s what some economists think will happen as robots and artificial intelligence increasingly become capable of performing human tasks. Of course, past technological upheavals created more jobs than they destroyed. But some labor experts argue that this time could be different: Technology is replacing human brains as well as brawn.
When politicians talk about jobs, they tend to focus on iconic, goods-producing industries, such as mining, steel production and auto making, that have traditionally been the hardest hit by global competition and technological progress. Lately, though, the loss of manufacturing jobs in the U.S. pales in comparison to the much larger losses in parts of the services sector.
Overall, services accounted for three-fourths of the job losses among more than 350 sectors of the private economy in the last year. That’s a big shift from previous decades, when goods-producing categories tended to suffer the most losses.”
Bloomberg used the retail industry as an example, noting that as customers increasingly purchased goods via the internet, department stores, which employ 25 times more workers than coal mining companies, are shedding workers at an accelerating rate. In the retail industry more broadly, average employment in the first four months of 2017 was down 26,800 from the same period a year earlier, against just 2,800 job losses in coal.
In retail and beyond, the modern services industry - which accounts for more than 70% of the US's economic output - is facing unprecedented challenges. Here’s a breakdown of some of the research cited in Bloomberg's analysis.
- The true extent of job losses could be much more severe than most workers expect. As Bloomberg notes, researchers at the University of Oxford estimate that nearly half of all US jobs may be at risk in the coming decades, with lower-paid occupations among the most vulnerable.
- In the U.K., the Bank of England estimates that about 15 million mostly service jobs—half the country’s total—could succumb to automation and widen the gap between rich and poor.
- A McKinsey Global Institute study of the labor force in 46 countries found that less than 5 percent of occupations could be fully automated using today's technology, but almost a third of tasks involved in 60 percent of occupations could be.
But if robots are truly taking over, mainstream academics would ask, then why haven’t we seen the attendent rise in productivity that one would expect from the increase in labor power?
While it's true that, in the past, innovation has led to job creation, it's foolish to believe that this trend will continue uninterrupted, especially as machines learn to perform increasingly high-level functions. As we’ve noted in the past, most of the new jobs that have been created are in low-wage, moderate-skill positions that cannot move the productivity needle much, causing the creation of new full-time jobs to stagnate.
But even if the academics are right and new high-skill jobs emerge to replace the ones that are being automated away, huge disruptions would still await. Large portions of the global workforce would still need retraining. And if work becomes a luxury, widespread joblessness and greater inequality could make it increasingly more difficult for the government to maintain social order.
* * *
To close out, here is a snapshot of the math that Cowen analyst Andrew Charles used to calculate the impact of McDonald’s “Big Mac ATMs” on the company’s minimum-wage workforce.
“MCD is cultivating a digital platform through mobile ordering and Experience of the Future (EOTF), an in-store technological overhaul most conspicuous through kiosk ordering and table delivery. Our analysis suggests efforts should bear fruit in 2018 with a combined 130 bps contribution to U.S. comps. We believe mobile ordering better supplements the drive-thru business where 70%+ of U.S. sales are transacted. In our view, MCD's differentiation lies in the operational enhancements of mobile ordering that includes curbside pick-up of orders in order to not disrupt the drive-thru.”
We are most excited for mobile ordering, Experience of the Future and the launch of fresh beef to help drive U.S. same store sales in 2018. We provide analysis for the latter three, which cumulatively we expect to contribute roughly 150 bps to U.S. same store sales in 2018, respectively. This gives us confidence to raise our 2018 U.S. same store sales forecast from 2% to 3%, in excess of Consensus Metrix’s 2.5%. Experience of the Future Features Lower ROI Than Mobile Order, But Offers Greater Potential Longer Term We are constructive on the use of guest facing technology for the restaurant industry. MCD’s longer-term U.S. story revolves around Experience of the Future (EOTF), a holistic operational and technological overhaul to the store base. MCD’s March 2017 investor meeting centered around the initiative with interactive displays. Perhaps the most conspicuous piece of Experience of the Future lies in digital kiosk ordering, which have seen success in International Lead Markets. Additionally, food ordered via the kiosk is delivered to the customer’s table. We believe EOTF better enhances the instore experience, which represents roughly 30% of domestic sales compared to mobile ordering, which allows customers to avoid leaving their cars.
Our ROI math suggests EOTF leads to a 9% cash/cash return in Year 1 in the 55% of domestic stores that do not require a store remodel, and 5% in the 45% of stores that require a remodel, which is a predecessor to implementing EOTF. Our math is premised on total costs of $150,000 for the Experience of the Future enhancement, and $700,000 of all-in costs when including EOTF as well as a store remodel. MCD has offered to pay 55% of the cost for Experience of the Future, in excess of the 40% the company contributed to the store remodel initiative beginning in 2010, for restaurants that commit to the program by the end of 2017. McDonald’s targets a high-teens return on incrementally invested capital (ROIIC, or Mcspeak for evaluating ROI), improving to the mid-20% range beginning in 2019. We believe EOTF’s ROI is captured over time as the sales lift does not dissolve as in the case of a traditional restaurant remodel. Rather, the lift should sustain as we expect consumers to increasingly embrace technological change. This is evidenced across concepts, such as Panera’s experience with 2.0, as well as McDonald’s own experience in Canada, where kiosks saw 12-13% sales mix in Year 1 and 27% in Year 2. We also note kiosk ordering will also likely lead to labor savings over time which should help boost ROIIC, but is unlikely for the foreseeable future.
In 2017, MCD expects to end the year with EOTF offered in 2,500 domestic locations from 500 at 2016-end. MCD targets much of domestic locations to feature EOTF by 2020, but has not given intermediary targets. The amount of stores adding EOTF depends on franchise reception to the initiative but we see positive indicators given our checks as well as the company’s disclosure that 90% of franchisees approved of the initiative after taking the same interactive tour that was given at the March 2017 investor day. We estimate 3,000 locations to add EOTF in 2018, which should lead to a 70 bps contribution to U.S. same store sales assuming an even cadence of restaurants adding the initiative over the course of the year. Further we assume the mix of stores adding EOTF in 2018 reflects the mix of overall stores needed to add EOTF, or 55% of stores that already have a remodel while 45% require a store remodel. McDonald’s has previously announced plans to remodel 650 restaurants in 2017, which we expect will also add EOTF.