I grew up steeped in the concepts of the “differential diagnosis” at the dinner table. Both of my parents were physicians – at a time when this was a true rarity – and they applied the concept to solving pretty much everything. Consider the observed evidence. Determine fact versus speculation. Generate a hypothesis. Narrow the array of possible answers by elimination. Loop again and again. Evaluate the answer you generate afterwards. This family meme generates a tendency to talk things through out loud – not great for cocktail party conversation, but pretty good for solving problems.

This hypothetico-deductive approach can be considered the basis for the scientific method, but the discipline can also be extended to process design and business efficiency. Well before computers, individuals looked at what they were doing and made incremental improvements based on observation, experimentation, and elimination of methods that didn’t work. In real life, of course, this process can be sloppy – how many times have we all said, “I guess that’s good enough?” At some point, your competitor went beyond good enough all the way to best. Bit by bit, agriculture and businesses did things a little better each year.

First up, of course, was just a better method. Then, tool design – a better plow or saw – generated stronger results. Later, came better machines. The Luddites famously objected to weaving machinery in the 19th century as their manual jobs were eliminated. A hundred years later, textile workers operating modern versions of the same machines object to people in other countries operating them more cheaply. The cycle of efficiency moves on.

Today, manufacturing is on the hot seat of various political debates. The generally accepted theme is that manufacturing in the U.S. is declining and that offshore manufacturing is the key culprit. It’s a story that makes sense; after all, we do see large amounts of imports. There are jobs that have been shifted offshore. Economic activity in places like China has exploded, and much of the growth has been achieved through the export of manufactured items. Many people are impacted and politicians circle to “do something about this.”

Well, back to checking the facts as part of our diagnosis. As shown in Figure 1, it turns out that after a severe decline around the “Great Recession” of 2008, total manufacturing output has recovered and remains pretty much on its long-term growth trend line that has been in place since the early 1900s.

Fig. #1
Industrial Production Index
01/1919 – 03/2017

Source: Board of Governors of the Federal Reserve System (US), Industrial Production Index [INDPRO], retrieved from FRED, Federal Reserve Bank of St. Louis; fred.stlouisfed.org/series/INDPRO, May 4, 2017.

(View a larger image here)

The key political issue, of course, is that the pace of job growth has not kept up with production. One estimate from researchers at Ball State University suggests that productivity changes – better efficiency – is the culprit. They state that, “Had we kept 2000-levels of productivity and applied them to 2010 levels of production, we would have required 20.9 million manufacturing workers. Instead, we employed only 12.1 million.” 1 That is a massive difference.

In fact, these same researchers estimate that “almost 88 percent of job losses in manufacturing in recent years can be attributable to productivity growth, and the long-term changes to manufacturing employment are mostly linked to the productivity of American factories.” Imports appear to account for around 13% of the changes in employment (while net exports and other factors like capital, both contributing to job growth, make up the rest to solve for 100).2

One might find it interesting if the politicians striving to “fix” employment from both right and left were working with the correct differential diagnosis. Are we treating for a sprain (trade), when it’s actually a fracture (efficiency)?

Where Does Automation Lead Us?
Over the next 20+ years, the real issue that the economy will be facing is how automation and efficiency impacts the non-manufacturing work force. As I look around the business world, including functions in our own operation, I observe that initially any rules-based job will be in jeopardy as technology improves. Analysts sit with an operational team; observe what takes time, what rules are used to determine an action, and what actions are taken. A little bit of code is then applied to eliminate the task while providing supervisors with alerts to override when the rules don’t fit. In office work, this generally results in lower growth in employment rather than elimination of jobs, but eventually that could occur, too.

Machines that work on their own with little supervision come next. Self-driving cars are fantastic in that we can all relate to what they need to do, and we are appropriately amazed with the possibility that soon we’ll be seeing them everywhere. My own two cents is that in the initial stages of use, there will be accidents and fatalities with self-driving cars. However, if the resulting plaintiffs don’t put the car companies out of business, the end result is likely to be a significant reduction in driving deaths and a lower cost of personal transportation.

On the employment side, I look forward to fleets of autonomous trucks crossing the country. These will, in effect, generate the efficiency of railroads without the need for fixed roadbeds. I expect better safety records – while acknowledging the excellent record out there with operators – and a significantly net lower cost of delivery without the need for drivers. But, that means that a portion of the estimated 3.5 million drivers and 5.2 million support personnel3 could be eliminated. Perhaps for a time, unions will prevail in requirements for having a driver present like they did with railroads and airlines, but in time, that, too, will likely disappear.

The concept of artificial intelligence adds some intellectual workers to the list of endangered species. Way back in the early 1970s, we would write elementary code that “learned” and extended itself. This was usually a form of guessing game where we would have the computer – a room-sized machine with the brain of a flip-phone – ask yes/no questions that branched towards a guess. When it couldn’t guess, it would ask the user to enter a new question that would generate the correct answer. These questions were then automatically added to the code.

This trivial example from nearly 50 years ago has been improved upon at the pace of Moore’s Law to the point that some researchers think that human and machine reasoning will converge over the next 15 years or so. Others think that it will never occur; but never is a very long time, indeed. Does this eliminate work? I certainly doubt that. But, even getting a little closer to the “singularity” will change employment radically.

We are all Luddites when it comes to eliminating or otherwise downgrading the responsibilities of our own jobs. A machine couldn’t do my job. Or couldn’t do it as well. Or shouldn’t do it. The strength of the argument declines as the end approaches.

So, as this next shift towards efficiency takes place, what will people end up doing? How will they make ends meet? Who will still be employed? What power will they have? Will we still use the same methods to measure productivity – dollars and cents? Will the “guaranteed basic income” currently being floated without a chance of enactment in many government circles become the standard?

Your Assignment, Should You Decide To Accept it…
Science/speculative fiction has dealt with ideas like this for as long as the genre has been in place. Take a few minutes to extract yourself from the current debates – mostly focused on the past – to speculate a little and suggest how things will evolve past 2050. Who holds power? What form of government exists? Who is still working in a traditional job that is around today? How are they paid? Have there been any revolutions and, if so, what did they look like? Send me your responses and I’ll go through them and use my judgment to see what warrants more discussion.

What types of jobs do you think will be replaced in the future by automation? Let us know by posting your thoughts in the Comments section of our blog.

Until the next Daily Pfennig® edition…

Onward and upward,

Frank Trotter
EVP & Chairman
EverBank Global Markets Group








All statements, comments and opinions expressed are solely those of the writer or speaker and are not the statements, comments or opinions of EverBank or of any of its affiliates, and are subject to change without notice. Due to the rapidly changing nature of currency and commodities markets, any statement, comment, or opinion may quickly become outdated. This is not a solicitation for the purchase or sale of any securities or options on securities or for the purchase or sale of a currency or any precious metal, and it does not constitute a recommendation to you or to any specific person of any particular action. EverBank, its officers and employees do not provide investment or other types of advice. All factual information has been obtained from sources that the writer or speaker believed to be reliable, but the accuracy, completeness, and interpretation of the factual information is not guaranteed and has not been independently verified. Not all products are right for everyone. You should conduct your own research and/or consult your advisor before making any purchases.

Click here for important additional disclosures.