Labor's Uncertain Future
Back in the stone age when I was growing up, we were taught that the automobile was responsible, directly or indirectly, for a large slice of the American labor force. Not just in assembly plants, but in production of steel, glass, rubber, and later, plastics and products made from petrochemicals. In addition to that, there were sales and service people, parts suppliers, warehousing and distribution operations. The operation of the vehicles required massive investments in oil exploration and refineries, which required more steel, and pipeline and rail distribution.
There have been some massive changes over the years in this key sector, which has long held pride of place in the U.S. manufacturing pecking order. First, import penetration began to affect production. But some of the foreign brands became so successful, they opened up factories here to service what eventually became their biggest market. Some of these factories even export finished cars back to their home countries. But a stunning thing has happened over the course of time, and it tells a story about the future of manufacturing in general and the future of the labor market as well.
Having come from a manufacturing background, I can tell you that the job of building or assembling any product involves a relentless pursuit for low costs, in terms of labor or materials. We've all seen how the rush to low-cost labor has sent an exodus of manufacturing capacity overseas. When I was in the apparel business, I witnessed a decades long progression of manufacturing facilities moving from the heart of Manhattan, to rural Pennsylvania, to the American South, to Central America and the Caribbean, to South Korea, Hong Kong (under British rule then,) Taiwan, and ultimately, the opening of China and the Indian subcontinent, where a great deal of our apparel comes from today. This decades process of global site-hopping represents an endless search for the cheapest production costs possible, and that is what manufacturing is all about. Manufacturing also made some of those nations rich, as it did our country: you rarely see apparel manufacturing in advanced economies like Taiwan or South Korea, and the rent is bit high in Hong Kong to site a cut-and-sew operation these days. So they've cast off their manufacturing base to some extent, just as we have, in favor of higher technology and finance.
But there is another factor in the manufacturing world that is causing a good deal of erosion in the need for human labor: automation. And we're seeing it across many industries, but the automobile angle caught my attention awhile ago when there was a fight to unionize the Volkswagen plant in Tennessee. It became something of a litmus test for the UAW's ability to penetrate the traditionally non-union South, and ultimately, they lost the vote by a small margin. But something about the story caught my eye. The plant produces about 100,000 Passats a year, with a maximum capacity of 150,000 units. The size of the assembly workforce was less than 1500 souls. That's all it takes in raw manpower to build 100,000 cars. The robots do the rest. Talk about a Pyrrhic victory, even if the UAW had won. Back in the 1970s, it took about 75 hours to build a car. Today, it's less than 24 hours.
Moreover, the robots have taken the tasks that required the highest skill set and craftsmanship, and therefore commanded the highest salaries. This too, is deliberate. The quest to “de-skill” the assembly process through automation is an integral component of manufacturing, and has always been so, since the days of the Luddites, who famously smashed the looms that came to commoditize their unique craft skills. The Industrial Revolution made a mockery of the Luddites, who have been looked upon with derision since those times. But it could be that the benefits of society from advancing automation may be reaching a kind of tipping point.
Take a look at this chart comparing auto output in units and the workforce needed to produce them. The relationship has essentially flipped.
Note that while output of autos and light trucks heads for an annualized rate of 18 million units (red line), compare the sheer number of workers required (blue line) to generate that level of output. What once took over 1.3 million workers to produce, now takes a bit over 900,000. And that's just for one segment of the industry. And you can see how this trend started back in the late 90s.
This trend has been quite active in many manufacturing sectors. But I wanted to focus on the automobile for several reasons. One, it's traditional role in American production, consumption patterns and wealth. Two, the possibility of further erosion of the industry due to new technologies, including the electric vehicle, whose mass adoption, should it happen, essentially obliterates dozens of industries that currently supply auto production, and self driving cars, which could not only put most of the auto service, repair and body shops out of business, but also obviate the need for a large percentage of spare parts needed. Together, these two factors could cause mass redundancy in the labor market. And it could happen quickly.
When I first started in the investment business, I had to make thousands of cold calls to business owners. One thing I couldn't help but notice was that every single travel agent I called had a disconnected phone. It took no time at all for internet services like Orbitz to completely vaporize an industry. Adoption of the new technology was quite rapid, and unforgiving.
You can't stop progress. But be mindful of the implications.