Automation and Technology Continue to Increase Living Standards

Transcript

Automation and technology offer the potential to solve many of the world's problems while improving our quality of life.

Just consider a few examples of the benefits of automation:

• Artificial intelligence (AI) will accelerate the drug discovery process from decades to months by testing the interactions of all known compounds.
• Driverless cars and trucks will improve traffic safety, eliminating traffic jams, saving hundreds of thousands of lives each year, and even reducing the cost of ground transportation.
• Robots will assist humans in providing eldercare to the aging members of society, starting with the Baby Boom generation.
• AI will enhance national security; for example, by using facial-recognition software to detect known terrorists and criminals in crowded places.

The applications are endless. However, according to pessimists, the negative consequences of technology are just as significant.

The primary fear that automation inspires in these people is that computers, robots, and AI will take over most of the work done by humans, leading to massive job losses, a wider income gap between highly-skilled and low-skilled workers, and ultimately the collapse of civilization.

What does history tell us? Every revolution in technology disrupts industries and displaces workers through the process of creative destruction. But inevitably, workers end up learning new skills and taking jobs in the new businesses that are built around the new technology.

For example, 100 years ago, one in three Americans worked on farms. Because food was so expensive to produce, the typical U.S. household spent 50 percent of its income on food.

Then automation - including technologies such as the tractor - made agriculture more productive. By 1960, less than one in ten Americans worked on farms. But this technological revolution did not lead to widespread unemployment.

Instead of working on farms, millions of Americans worked in factories and offices. And because they needed to spend less money on food, they could buy the products that came flowing out of those factories - everything from toothpaste to televisions.

Then the economy went through another shift, as machines and robots gradually performed more and more of the work done in factories.

Between 1960 and 2014, according to the U.S. Bureau of Labor Statistics, the proportion of Americans working in the manufacturing sector plunged from 28.4 percent to 8.8 percent. Again, the displaced workers landed on their feet - this time in the service sector.

Millions of new jobs were created, and consumers enjoyed both cheaper manufactured goods and new services. Despite this historical pattern, some economists claim that "it's different this time," meaning that workers will be left permanently jobless as machines and AI replace humans in more and more roles.

For example, Vivek Wadhwa of Stanford University recently opined, "Advancing technologies will cause so much disruption to almost every industry that entire professions will disappear. And then, in about fifteen to twenty years from now, we will be facing a jobless future, in which most jobs are done by machines. . . ."

Similarly, MIT professor Andrew McAfee told CBS News 60 Minutes that "When I see what computers and robots can do right now, I project that forward for two, three more generations, I think we're going to find ourselves in a world where the work as we currently think about it is largely done by machines."

McAfee's colleague at MIT, Professor Erik Brynjolfson, told the same interviewer, "[Technological advances are] always destroying jobs. But right now the pace is accelerating. It's faster, we think, than ever before in history. So as a consequence, we are not creating jobs at the same pace that we need to."

However, as James Sherk and Lindsey Burke of the Heritage Foundation observed in a recent report, such claims are misplaced. Why? Consider these two facts: Productivity growth has slowed down over the past decade.

In the most recent year for which data are available, labor productivity actually dropped by 0.1 percent. If automation was truly replacing human workers, we would expect productivity growth to speed up.

That's because companies would be producing the same or more goods with fewer people - but that's not what the statistics show. The number of lower-skilled workers - those who are supposedly most vulnerable to being replaced by machines - has increased in the workforce, and their wages have gone up.

If automation was making these workers obsolete, we would expect their employment and wages to decrease. But compensation for workers in the bottom quintile of the income distribution went up 58 percent from 1979 to 2011, the most recent year for which data are available.

Only the top quintile grew faster, while the three middle quintiles saw their compensation grow at least 25 percent slower than that of the lowest-skilled workers.

On top of that, keep in mind that automation of routine jobs leads to lower costs for businesses, which results in lower prices for consumers. So, just as in past waves of automation, everyone - low-skilled workers and up - benefits from greater spending power.

And, because of that greater spending power, demand is increasing for products and services in other parts of the economy, where human workers are needed to meet that demand. It's easy to see why people can make the mistaken assumption that the increase in automation will put humans out of work.

The "lump of labor" view of the economy has been discredited, but it still influences people's thinking. This view holds that there is only a certain amount of work that the economy needs to have done.

Therefore, if machines take over ten million jobs from one decade to the next, that would mean that ten million human workers would become unemployed. Fortunately, that's not the way the real world works.

Because of ever-expanding consumer demand for new offerings, new jobs are constantly being created throughout the economy. As Sherk and Burke explain, "Automation does reduce the human labor needed to produce particular goods and services, but it also reduces production costs. Competition forces firms to pass these savings on to their customers through lower prices. These lower prices lead consumers to buy more of the now less-expensive product and leave them with more money to spend elsewhere, increasing the demand for labor in those sectors of the economy. The amount of work in the economy expands to use the available labor supply."

In fact, they found that average U.S. hourly labor productivity went up by 108 percent from 1973 to 2014. At the same time, average hourly inflation-adjusted compensation of U.S. workers went up by 85 percent.

So technological advances that made higher productivity possible didn't cause wages to fall. Instead, wages went up because demand went up, thanks to lower prices that were driven by increasing automation.

It's not just that a shrinking number of workers enjoyed these higher wages. The employment-to-population ratio for workers between the ages of 25 and 54 actually went up from 1973 to 2014. What's behind these numbers?

How can we live in a world where automation is increasing, yet more humans are working, and their compensation is increasing? To understand what's happening, we have to understand that machines, robots, and computers are still very limited.

They are increasingly adept at mastering routine work that doesn't require any ability to make a judgment, such as drilling a hole on an assembly line. But they can't keep up with humans on non-routine tasks, which are plentiful in the workplace.

According to MIT economist David Autor, a non-routine task is anything that a computer programmer hasn't explicitly coded the machine to do. Everything that a robot or computer does is the result of it following specific instructions, and there are simply too many situations that humans respond to easily that machines can't handle.

For example, Autor writes, "The challenges to substituting machines for workers in tasks requiring flexibility, judgment, and common sense remain immense. . . . Fundamentally, computers follow procedures meticulously laid out by programmers. For a computer to accomplish a task, a programmer must first fully understand the sequence of steps required to perform that task, and then must write a program that, in effect, causes the machine to precisely simulate these steps. . . ."

"Large online retailers, such as Amazon.com, Zappos.com, and Staples.com, operate systems of warehouses that stock, pack, and ship thousands of varieties of non-homogenous goods directly to consumers and businesses. These warehouses employ legions of dexterous, athletic 'pickers,' who run and climb through shelves of typically non-air conditioned warehouses to locate, collect, box, label, and ship goods to purchasers. There is at present no technologically viable or cost-effective robotic facsimile for these human pickers. The job's steep requirements for flexibility, object recognition, physical dexterity, and fine motor coordination are too formidable."

However, robots can increase the productivity of the human workers by taking on the routine work that can be automated. In the case of the warehouses, robots can move pallets of products to the workers, who then pick the products before the robots move the pallets away.

In countless jobs, the work performed is a mix of both routine and non-routine work. For example, a product designer might create an innovative new product that is beyond the capabilities of any machine.

Yet her productivity is enhanced by tools like Google to search for information, CAD to design the product, Excel to create a spreadsheet, and PowerPoint to present her idea.

Looking ahead, we offer the following forecasts: First, while automation will not cause large numbers of net job losses, some occupations will be automated, forcing human workers to be retrained for other careers.

As noted earlier, the occupations that are ripe for disruption are those that consist of performing routine work that a machine can be programmed to do.

Sherk and Burke cite the examples of hotel employees, fast-food restaurant workers, and medical researchers. In Japan, a new hotel will be almost entirely automated, with robots checking in guests at the front desk, moving their luggage to their rooms, and cleaning the rooms after checkout.

Meanwhile, a company called Momentum Machines has introduced the Alpha hamburger-making machine, which can produce six hamburgers in a minute on a robotic assembly line.

IBM's Watson supercomputer compared the findings from more than 70,000 medical research papers to identify six proteins deserving of further investigation; that same amount of work would have taken a single scientist 38 years to complete.

Second, companies will gain competitive advantage not by replacing people with technology, but by integrating the work of humans and machines. Non-routine work will continue to resist attempts at automation, because machines are not flexible enough to adapt to changes in their environment.

However, human workers will become increasingly productive as more of the routine work that wastes their time is delegated to machines. Third, the increasing automation of routine work and routine tasks that complement non-routine jobs, will lead to a new era of American prosperity.

Automation will continue to lower the costs of goods and services, even as wages increase and net employment is largely unaffected by the increasing adoption of technology.

The result will be a virtuous cycle in which workers have more money to spend on products and services that cost less and less; this will elevate living standards for workers of all income levels and fuel the growth of the overall economy.

Fourth, policymakers will need to eliminate excessive regulations in order to allow for the faster redeployment of lower-skilled Americans whose routine-work jobs will be lost to automation.

As Sherk and Burke point out, barbers, florists, interior designers, bartenders, cosmetologists, and drywall installers currently need to be licensed after undergoing training that can last more than a year for some occupations.

Regulations such as these can keep people from making the transition to non-routine jobs where they can earn a living while benefiting from the lower prices that the automation of their former jobs will bring.

Comments