NBC’s cable business channel, CNBC, once had a program on Frito Lay, the leading potato chip provider in the United States. Apparently, there is a handful of Frito Lay processing plants on the continent. The show visited the southeast area plant that is staffed by just a few people. The plant is fully automated. One person inspects large batches of potato trucks before their potatoes are accepted for processing. There is a small group supporting the machines and some more getting the product out the door. The point is that just a handful of people can feed almost about a quarter of our population’s needs for potato chips. It is a pristine example of automation in our society, and the appropriate economic question to ask of these dedicated men and women at the Frito Lay plant is how many potato chips can they eat?
Before we talk about automation in the modern day, lets trace its development from its beginnings and that starts with Henry Ford.
Ford was a pioneer. He developed the beginning of modern day mass production processes. He implemented the moving assembly line that brought the product to each worker, each of whom did a repetitive task on the product. They did their respective tasks over and over again until they were bored, and sore, and achy.
Ford believed in his workers. He developed a town with houses they could live in that was a short distances to the plant. A famous story is a worker who spent all day on his back attaching some part of the drive train to the chassis of each car. The worker would come home day after day and stare into the fire in his fireplace and just try to recover. He did nothing else. He was physically exhausted. What Ford did not know is that each worker became a machine when in reality they were humans. Walter Reuther would realize this fact and organized the United Auto Workers. This led to a lot of violence on the UAW’s path to full recognition in the 1940’s when Henry Ford II took over.
Ford believed in Taylorism, or what we know as Scientific Management, which is defined as one best way to perform each task. The process that was most efficient was also the most profitable. Again, the worker became like a machine through the eyes of Ford and his application of Scientific Management. Ford also believed in sourcing parts to companies that were in the neighborhood of the plant. The beginnings of just-in-time parts inventory had its intellectual foundations at Henry Ford’s production plants in Dearborn Michigan.
But Ford also realized something that his peers at the other car companies did not. The workers had to be paid enough to afford what it is that they were building. He paid each worker $5 a day, which was a lot for a laborer in those days. Ford workers could afford to buy Ford cars. This is a major economic reality. One must have economic consumption in order to have economic production. Ford could not get rich unless his workers and others bought his cars. Ford’s business model provided for this. Hence, each worker could afford to buy a least one car.
The Japanese studied Henry Ford’s production processes when they began manufacturing in the 1950’s. They took the organization practices of Ford and his protégé, William Knudsen, and applied them to their own situation. Soon they advanced local suppliers into JIT. They also noted that humans producing things was problematic, especially when it came to quality. So they adopted the quality management philosophy of W. Edwards Deming in the 1980’s, and used sensors and computers to collect statistical information on quality parameters in their cars and other products they manufactured. This allowed them to identify problems in their products that were in the most need of fixing and they fixed them. During the 80’s, people would marvel at the fit and finish of Japanese cars and they cursed the problems with American ones. Fortunately, American car companies have become leaders in production techniques in the last couple of decades.
In 1969 American Machine and Foundry purchased Harley-Davidson, America’s premiere motorcycle manufacturer. They streamlined production, failed to observed quality, and developed a reputation for bad bikes. Sales plummeted. In 1981, the company was sold to a group of 13 investors that included Willies G. Davidson, a grandson of one of the H-D co-founders. JIT was introduced, inventory was strictly observed, and robots and computer driven processes were introduced. Quality was improved and H-D was profitable again. Prior to the buyout, fifty plus people machined the transmission housings with quality problems in a substantial portion of the product. After the buyout, the housings were manufactured by 3 people using computerized machines with quality problems in less that 0.01% of the product.
Now we come to a Frito Lay plant that is run by a handful of people who can supply the needs of at least 20% of the country. Can they consume all that they produce? The Ford workers could consume a good chunk of what they produce, but I am not so sure of the Frito Lay workers. Their collective salaries couldn’t even buy a small chunk of the product. People need to produce in order to consume. Computers and robots can’t consume what they produce. This disconnect has tremendous economic implications.
CKE President Andrew Puzder has been appointed to the post of Labor Secretary. CKE runs the Carl’s Jr. and Hardy’s fast food restaurants. Apparently, Puzder does not believe in a $15 an hour minimum wage, and in order to cope with the coming wage standard, Puzder was moving CKE towards automating many of food preparation and serving functions in its restaurants. In short, CKE is adapting to higher labor costs with higher capital investment. In today’s economic culture, this is understandable. It also begs the question that is can those most likely to consume a CKE product, namely CKE’s own employees, more easily or less easily afford to buy that product?
Dick Durbin is the current senior Senator from Illinois in the US Senate. He was preceded by the legendary Paul Simon, who was preceded by Charles Percy, who was preceded by Paul Douglas, a very notable individual outside the Senate. He enlisted in the US Marines at the age of 50 during the Korean War, one of the oldest recruits in its history, and rose from a private to the rank of Lieutenant Colonel, eventually earning a purple heart. But the experience I want to discuss is his contributions to economics. Paul Douglas was an economics professor for years at the University of Chicago, and a good one too, capping his career as President of the American Economic Association in 1947.
In 1928, Paul Douglas and Charles Cobb published an article that proposed a production function that is widely used today. The Cobb-Douglas Production Function reads as follows.
Y = ALaK1-a
The statement reads as follows. Total production is equal to a scalar factor (A) times labor (L) to an exponent of less that one (a) times capital (K) to a complementary exponent (1-a). What is important about this relationship is that capital and labor can be viewed on a continuum. The more labor you use to produce an item, the less capital you need. The reverse is also true. The more capital you use in production, the less labor you need. Economists often graph this function with labor placed on the horizontal axis (X) and capital on the vertical (Y). If you can imagine an inverse quarter circle hugging the two axes at their end points, then you can see how an increase in capital reduces the use of labor graphically.
Even though there are other production equations that economists use, the Cobb-Douglas function is the most commonly used function. If one draws a Cobb-Douglas production graph, it is important to take the labor number from the labor axis into a separate calculation. Total labor hours times average wage gives total consumption from the everyday workforce. When you shrink total labor hours due to automation, you shrink consumption, which eventually shrinks production, even if it is automated.
This whole explanation brings us back to the basic question at the top of the post. When firms like Frito Lay and CKE automate their production, they inherently reduce the amount of consumption that the public can support. When Frito Lay automates their plant, then other potato chip providers will automate theirs since they need to match Frito Lay in cost structures. If CKE automates, then other fast food chains will also automate. The loss of purchasing power and consumption at the lower levels of our workforce will continue to widen the wealth effects between lower and upper level strata in our economy at an accelerated rate.
In essence, companies are shooting themselves in the foot when they automate at the expense of local jobs. In the case of Frito Lay, a few people can provide enough product for a quarter of the US population. If the few people who operate the plant could physically eat all the product they make then we could visualize sufficient trades of chips for televisions, iPhones, computers, cars, houses, and other baskets of products. But there are not enough wages within this small group of high earners to support these product baskets when we factor in the larger population. If all our industries replicated Frito Lay and CKE, then we as a population will be basically poor. The excess product from all the automated plants will have to be exported or be scrapped because we, as a population, could not afford the product.
There are lots of lost plants that were moved overseas to take advantage of low-cost foreign labor. If we are repatriating these plants because we can produce an equal or better product due to automation, then we can add a handful of new employees for each new automated plant we place online thereby marginally increasing our workforce each time. But if we are reducing our existing workforce with automated plants, then we are reducing our purchasing power, our consumption, and eventually the markets of those manufacturers that want us to buy their products. In short, they could be shooting themselves in the foot through all the automation.
This is just one issue with the new labor economics that Trump is proposing. Additional issues will be discussed in future posts, and these issues cannot be isolated from one another. Some posts will discuss their independencies with one another. For now, it is fair to conclude that not all automation is good, and that we have to understand where the labor will come from to buy all the goods that the automated plants produce.