telegraph.co.uk

Technological Dilemma

Thomas Black of Bloomberg recently reported on two federal regulatory initiatives that appear to be going in opposite directions:

One arm of the U.S. Department of Transportation is recommending two-person crews be required for freight trains as another plans to spend billions to help develop driverless technology for long-haul trucks. [1]

Railroads in the U.S. typically operate freight trains with a two-person crew: a conductor, who is in charge of the train (paperwork, schedule, etc.), and the engineer.  In the days of cabooses—a common sight that passed into history in my lifetime—the conductor and other crew travelled in the caboose, which served as the conductor’s office and quarters for additional crew used before technology and economics eliminated those positions. Now, the only reminder of the caboose is the blinking red light attached to the end of a freight train.

The argument today surrounds maintaining the two-person crew or dropping down to the engineer as sole crew member.  The railroad unions argue for the two-person crew on the grounds of safety, and the federal regulators appear to agree—for the moment, at least. The railroads say they are spending huge amounts on technology by installing (by Congressional mandate) automatic train control systems to prevent derailments and collisions, but the system is not fully implemented.   

Driverless vehicles

Turning from rails to roads, we are all aware of the strides being made in driverless vehicles, which may seem like a fulfillment of Popular Science prophecies in the pre-computer 1950s.

It seems only a short time ago that my brother, an engineer with Ford at the time, visited us while doing some test driving of collision-avoidance systems.  The Ford Explorer he was driving was jam-packed with test instruments, computers, and other equipment.  The obstacles to a working system seemed daunting. Now, cars are available with a host of technology—cameras, self-parking, collision-avoidance, etc.—an incredible leap in such a short time.  

Most of these advances either assist the driver by providing more information for safe travel, or they actually take over control of the vehicle. So why not driverless cars?  Is it not just an extension of the same technology trajectory? But, driverless semis?

While the idea of travelling along the truck-saturated interstates full of driverless trucks can be unnerving, my first thought was “what will happen to all of the truck drivers displaced by automation?” 

Will drivers still be needed at some point along the way?  How big and what kind of workforce will be needed to manage and maintain fleets of driverless trucks? Is the primary motivation safety and efficiency, or economics?

Technology and the workforce

According to truckinfo.net, there are 3.5-million professional truck drivers in the United States, with a total of 8.9-million people employed by the trucking industry [2].  Of the 15.5 million trucks in the U.S., 2 million are tractor-trailers.

Those 3.5 million drivers, represent about 2% of the U.S. workforce of 157 million people (September 2015, according to Dept. of Labor statistics), nearly 6% for the entire trucking industry.  Complicating any analysis, however, is the fact that labor force participation reached a 38-year low of 62.4% at the same time, with 95 million not counted in the labor force. [3]

The participation rate had risen steadily from 59% in the early 1960s to 67% in 1999, when it leveled off and then began a precipitous and steady decline beginning in 2009.  Since changes in technology have been influencing the labor force for at least two centuries, the reasons for this trend—and especially the most recent downturn—seem to go far beyond technology.   

Technology has radically transformed many jobs and entire industries over history, particularly since the Industrial Revolution. In transportation technology alone, the geography of America has been transformed by the dramatic changes from river transport to railroads, to highways, to airways, and now the impact of the more ethereal Internet on communications and commerce (which relies on transportation).    

While each new technology was transformative, it did not necessarily eliminate the old.  Yes, horses were replaced by gasoline engines, which had huge implications for everything associated with horses,  shifting instead to the manufacture of engines. But, the makers of horse-drawn carriages more easily transitioned to supply parts for the new horseless carriages. 

So, how worried should we be that automation represents the dark side of employment—in transportation or any other industry?   Is automation fundamentally different from any of the previous transitions where job losses in one area were replaced by gains in others, even the formation of entirely new industries?

When I did a search on “automation and job loss” the top results were all very current—and the “research” and opinions were split between alarm and caution.  Following are brief excerpts of each post. Links to the full source article will be found in the Notes at the end.

Raising alarms – Three of the entries predicted serious job loss because of automation:

“Rise of the Robots Will Eliminate More Than 5 Million Jobs” by Jill Ward, Bloomberg, January 18, 2016, which reported on a World Economic Forum (WEF) report on automation and job loss:

Over five million jobs will be lost by 2020 as a result of developments in genetics, artificial intelligence, robotics and other technological change, according to World Economic Forum research.

About 7 million jobs will be lost and 2 million gained as a result of technological change in 15 major developed and emerging economies, WEF founder Klaus Schwab and managing board member Richard Samans said in "The Future of Jobs." The findings are taken from a survey of 15 economies covering about 1.9 billion workers, or about 65 percent of the world’s total workforce. [4]

Compare this story with Tim Worstall’s analysis of the WEF report on Forbes the same day, which appears below.  

“White House Predicts Massive Job Losses to Robots” by Nate Church, Brietbart, February 23, 2016

The [White House] report suggests an over 80% chance that jobs paying less than $20 per hour will eventually succumb to the cybernetic revolution. Jobs in the $20 to $40 look to be cut by about one-third, but positions at the high end should see less than 5% losses to automation. The numbers are staggering and, if true, represent the biggest shift in labor that the modern world has ever seen.

... It’s tempting to suppose that these are problems for another day, that the future is perpetually uncertain. But this future is standing on our doorstep, and we are only a few steps away from being forced to answer. We’ll need a workable solution, before large parts of humanity are simply rendered obsolete. [5]

“Tennessee Prepares for Loss of 1.4 Million Jobs to Automation” – blog post on vdare.com, April 6, 2016:

While our hidebound representatives in Washington DC continue to ignore the technological revolution increasing in the nation’s workplaces, the political leaders in Tennessee are stepping up. The state government recognized the problem, studied it and finally created a policy response in the form of an educational program. The nation’s leaders should do half as well.

[After describing educational efforts]...Still, the education solution may be only temporary in the coming techno-future as smart machines advance in abilities. Tech entrepreneur Vivek Wadhwa observed in a 2014 Washington Post piece: “. . .at best we have another 10 to 15 years in which there is a role for humans”  [The blog contains links to the Tennessee study, the Wadhwa article and other references]. [6]

The Other Side of the Coin

Compare those predictions with others that either debunk concerns for job loss due to automation or recognize the threat, but temper it with a broader, historical view.

“The Rise of Job-Killing Automation? Not So Fast” by Thomas Davenport, The Wall Street Journal, August 12, 2015, which stresses uncertainty about making predictions:

One key reason for the uncertainty is that as new technological capabilities develop, only certain aspects of jobs are replaced by technology. Substitution of smart machines for human labor works at the task level rather than at the job level. Technology takes over some things that humans do in a particular job, but not others.

... I am not saying that you should wear a T-shirt saying “Keep Calm and Forget Watson.” Automation and cognitive technologies are real and will be very influential in the job market. But predictions of loss numbers for individual jobs in 2025 are about as accurate as predictions of individual stock prices in 2025. Be concerned and be prepared, but don’t believe everything you read. [7]

“The Future of Automation—and Your Job” by Stephen Gold, President and CEO, Manufacturers Alliance for Productivity and Innovation (MAPI), Industry Week, January 16, 2016, in which he acknowledges a fundamental difference from previous changes because of the role of artificial intelligence:

Concerns about technology replacing workers isn’t new. Over the past 200 years, mechanical looms replaced artisans, electric streetlights displaced lamplighters, cars replaced horses (thus limiting the need for blacksmiths) and personal computers diminished the need for typists. Technological advances have long taken a toll on specific professions and industries. They also raise our overall living standards.

The difference today is that digitization is dramatically reducing transaction costs and helping spread the changes farther and faster than before. Moreover, a new kind of automation—artificial intelligence—puts different kinds of jobs at risk than the mechanical automation of previous eras. Particularly affected are tasks that require cognitive skills, many of which are found in middle-skill jobs. [8]

“WEF's Davos Report On Robots, Automation and Job Loss: A Trivial Result Of No Matter At All” by Tim Worstall, Forbes, January 18, 2016.  Following his analysis of a Reuters report that suggests dire projections in a World Economic Forum (WEF) report and the report itself, Worstall concludes:

That is, technological change happens during, because of, through, that turmoil of quits and fires and rehires [the normal churning of the job market described by Worstall]. And the WEF is telling us, in those very serious tones, that we should be worried about a 0.3% difference in a 100% flow? That’s preposterous. The truth of the matter being that they’re just not modeling the process as that flow at all. They’re looking at the static numbers and not considering in the least how the process actually comes about.

Which leaves us with what we need to do about this. The answer being nothing, nothing at all. The jobs market in general will have no problem whatsoever in dealing with an extra 0.3% of the technologically displaced over a 5 year period. It’s simply a trivial addition to the well understood process by which technological change happens anyway. [9]

What Should We Conclude?

As something of a futurist myself, I tend to watch and predict trends rather than quantitative outcomes. Too often, it seems, those who wish to be specific in their claims of doom or promise, have an agenda that must be considered.  We must also assume that unintended consequences can potentially derail overly optimistic promises, while intervening and more complex circumstances can mitigate dire predictions.

When looking at automation and job loss, my own concern is for the quality of work. Some changes eliminate tedious, menial, or dangerous work that is not good for the worker or the society in the long run.  Many jobs today have been improved because of attention to working conditions and a view of employees as an asset.  It does the worker and society no good when the elimination of any job forces workers into lesser positions or removes them from labor participation entirely.  

There is a relationship, I believe, between good work and a good society.  Change is inevitable. While it can be extremely difficult, we must anticipate change and greet it with enthusiasm and creativity. 

NOTES:

1. “Why Will the Feds Let Robots Drive Trucks but Not Trains?” by Peter Black, Bloomberg, June 30, 2016.
2. Trucking statistics, truckinfo.net
3. Labor statistics, October 2015, CNS News report
4. “Rise of the Robots Will Eliminate More Than 5 Million Jobs” – Bloomberg
5.  “White House Predicts Massive Job Losses to Robots” – Breitbart
   also see “White House Predicts Robots May Take Over Many Jobs That Pay $20 Per Hour” by Shahian Nasirpour on The Huffington Post, February 24, 2016
6. “Tennessee Prepares for Loss of 1.4 Million Jobs to Automation” – vdare.com
7. “The Rise of Job-Killing Automation? Not So Fast” – Wall Street Journal
8. “The Future of Automation—and Your Job” – Industry Week
9.  “WEF's Davos Report On Robots, Automation And Job Loss: A Trivial Result Of No Matter At All” – Forbes

See other InfoMatters blogs related to this topic:
Life, Liberty, and...  - July 4, 2016 – makes reference to the meaning of “the pursuit of happiness” in the Declaration of Independence and its application to the workplace.
Unintended Consequences, Part 1 – November 5, 2015 – the first part looks at consequences that make matters worse than the problem they were meant to resolve.
Unintended Consequences, Part 2 – November 5, 2015 – the second part looks at unexpected benefits.