Tariffs may kill more U.S. manufacturing jobs than they save. And it might be a good thing
Regardless of whether you’re a proponent or not, automation is happening, and it’s happening fast. According to a recent International Federation of Robotics report, by the end of 2016 there were over 1 million manufacturing robots operating in Asia, and this year that number will exceed the number of Foxconn employees for the first time. By 2019, the IFR predicts almost 40% of the global supply of robots will be installed in China, likely due to a Chinese government-supported initiative to become a global leader in automation.
The story in the United States is less clear. While 7.3 million manufacturing jobs have been lost over the past thirty years, productivity is at an all-time high. Sounds good, but the nuanced point is that most of the productivity gains have been in the electronics and computer industry—not broadly across manufacturing in general. It’s not difficult to see why: ten years ago, the iPhone hadn’t even been released. Since then, the “mobile revolution” has been underway, causing a surge in demand for high volume products like smartphones (made in Asia) and the servers, chips, and technologies needed to enable them (often made in the United States).
China has adopted policies with the intention to become a global leader in automation. What policies is the United States considering? The ones making the most headlines these days are economic policies to tax imports—with the intention of incentivizing the reshoring of manufacturing jobs.
The question no one is asking is: what will those policies do to American manufacturing automation?
The relationship between wages and automation
Interestingly enough, outside of piece-part manufacturing, semiconductor, and PCB assembly, most of these consumer electronics products are assembled by hands. Millions of human hands. Mass market consumer electronics are complicated products, usually super small and tightly packed with custom parts, that will only be made for three years or less. While it’s possible to build robots that can achieve the micro-manipulation needed to build these products, so far the calculation for Return on Investment (ROI) hasn’t checked out. These robots need to compete in a global race to the bottom with human labor, where today’s low water mark for consumer electronics of reasonable quality is about $3 an hour.
The ROI equation creates a delicate balance—even small perturbations can dramatically affect what starts to make financial sense and can change the industry as a whole. For example, a smartphone or similar product usually ships for three years—what if it was in production for longer? My favorite example of this is the Steam controller, where the final assembly has been fully automated in a factory in the Midwest. What if the complexity of what needed to be done was reduced significantly? Some portions of the assembly line, such as the “testing and packing” part of the final assembly process, can already be automated, and Foxconn is working on it. These perturbations have created interesting opportunities for automation before mass roll-out has made financial sense.
The elephant in the room
Recently, there has been an upswell of public discussion around “bringing the manufacturing jobs back” to the United States and how this should best be done. But what would unnatural market policies like import tariffs end up doing to the ROI calculation for automation?
First, we have to understand the economics at play in the decision of where to set up the workshop. The most obvious one is wages. Air conditioner operators in the United States make $23 per hour, while the same operator makes $4 per hour in Mexico or $3 per hour in China. A second consideration is supply chain proximity: where are the sub-components for the final product being made? This matters because the closer your supply chain is to your final assembly factory, the less inventory you need and the faster you will be able to react to swings in demand. Apple’s supply chain is famously short, only a couple of days—giving them a huge competitive advantage over their competitors whose supply chains are three to four times longer. The third consideration is shipping costs between the factory and the consumer, which partly depend on how heavy and bulky the product itself is. Small products can be air shipped for a couple of dollars and only a day in transit. Boat shipping is cheaper, but takes longer and is less reliable.
These factors are combined in a kind of calculus to determine the “best” place to manufacture a product. While a 20% tariff seems like nothing when you’re comparing $4 to $23 per hour—it can throw a wrench into the whole equation. If Mexico was formerly the “best”, tariffs may make the United States seem more appealing—but unfortunately, it is unlikely to do so at the benefit of the American worker. When coupled with the promise of producing the same goods for less at higher yields, automation equipment that isn’t able to compete with a Mexican operator becomes economically viable compared to an American one.
It’s time to embrace automation
In the past, mechanization has created new industries and opportunities that couldn’t be imagined previously — and I choose to believe the same will be true in manufacturing.
Ultimately, the public should be supportive of such advancements—not just because of the silliness of holding back just to hold back (as Milton Friedman quipped, don’t make men use spoons instead of shovels)—but because there is much to be gained through automation.
First, automation improves product consistency and quality, reducing scrap rates significantly. The amount of non-recyclable scrap and waste produced in the making of one smartphone is mind-boggling once scaled to the 5.5B ever made, filling every professional football stadium in the United States ten times over. As an engineer at Apple, I once calculated the environmental impact of increasing the length of a release liner pull tab by an extra millimeter—it won’t matter if I recycle my bottles and cans for the rest of my life, I’ll never make up for it.
Second, and most importantly, automation keeps the United States competitive on the global market. We too will be able to produce higher quality products, at lower cost, enabling larger markets or greater access for more consumers. In the past, mechanization has created new industries and opportunities that couldn’t be imagined previously—and I choose to believe the same will be true in manufacturing.
One final thought
Keeping the global automation playing field level will be important as the post-automation stage becomes more clear. I believe this next stage will be about automation software and intelligence. It will harness the same technology that determines what movies Netflix recommends to instead suggest improvements to the process or product. This intelligence will drive the next wave of manufacturing efficiency and productivity gains. If we embrace automation here at home, we will be positioned to ride this next wave. If we don’t, we may never catch up.