Rise of the Robots: Technology and the Threat of a Jobless Future (34 page)

The problem is that making that happen requires dramatically raising household incomes, as well as addressing the issues that have caused the savings rate to soar. Initiatives such as improving the pension and health care systems may help somewhat by reducing the financial risks faced by households. The Chinese central bank has also recently announced plans to relax regulations that hold down the interest rate paid on savings accounts. This might turn out to be a dual-edged sword, on the one hand raising the income going to households but on the other further increasing the incentive to save. Allowing deposit rates to rise could also threaten the solvency of many Chinese banks, which now profit from artificially low interest rates.
³⁵
Some factors behind the Chinese propensity to save may be very hard to address. Economists Shang-Jin Wei and Xiaobo Zhang have proposed that the high saving rate may be attributable to the sex imbalance resulting from China’s one-child policy. Because women are scarce, the marriage market is very competitive, and men often have to accumulate substantial wealth or own a home in order to attract a potential spouse.
³⁶
It is also quite possible that a strong desire to save is simply an integral aspect of Chinese culture.

It’s often remarked that China faces the danger of growing old before it grows rich, but what I think is less generally acknowledged is that China is in a race not just with demographics but also with technology. As we saw in
Chapter 1
, Chinese factories are already moving aggressively to introduce robots and automation. Some factories are reshoring to advanced countries or moving to even lower-wage countries like Vietnam. A look back at
Figure 2.8
in
Chapter 2
shows clearly that advancing technology resulted in a relentless sixty-year collapse in American manufacturing employment. It’s
inevitable that China must ultimately follow essentially the same path, and it’s quite possible that the decline in factory employment may turn out to be even more rapid than in the United States. While automation in American factories progressed only as fast as the new technology could be invented, China’s manufacturing sector can, in many cases, simply import leading-edge technology from abroad.

In order to negotiate this transition without a surge in unemployment, China will have to employ an ever-increasing fraction of its workforce in the service sector. However, the typical path followed by advanced nations has been to first become wealthy on the basis of a strong manufacturing sector and then make the transition to a service economy. As incomes rise, households typically spend a larger fraction of their incomes on services, thereby helping to create jobs outside the factory sector. The United States had the luxury of building a strong middle class during its “Goldilocks” period following World War II, when technology was progressing rapidly, but still fell far short of substituting completely for workers. China is faced with performing a similar feat in the robotic age—when machines and software will increasingly threaten jobs not just in manufacturing but also in the service sector itself.

Even if China does succeed in rebalancing its economy toward domestic consumption, it seems optimistic to expect that the country’s consumer markets will be fully open to foreign companies. In the United States, the financial and business elite profited enormously from globalization; the most politically influential sector of society had a powerful incentive to keep imports flowing. In China, the situation is quite different. The country’s elite are more often than not affiliated directly with the government, and their primary concern is keeping the regime in power. The specter of mass unemployment and social unrest is perhaps their greatest fear. There is little doubt that they would choose to implement overtly protectionist policies if faced with that prospect.

The challenges faced by China are even more daunting for poorer countries, which are much further behind in the race against
technology. As even the most labor-intensive areas of manufacturing begin to incorporate more automation, the historical path to prosperity may be poised to largely evaporate for these nations. According to one study, about 22 million factory jobs disappeared worldwide between 1995 and 2002. Over the same seven-year period, manufacturing output increased 30 percent.
³⁷
It is not at all clear how the poorest countries in Asia and Africa will manage to dramatically improve their prospects in a world that no longer needs untold millions of low-wage factory workers.

A
S ADVANCING TECHNOLOGY CONTINUES
to drive inequality in both income and consumption, it is poised to eventually undermine the vibrant and broad-based market demand that is essential for continued prosperity. Consumer markets play a critical role not just in supporting current economic activity but also in advancing the overall process of innovation. While individuals or teams generate new ideas, it is ultimately consumer markets that create the incentive for innovation. Consumers also determine which new ideas succeed—and which are destined to fail. This “wisdom of crowds” function is essential to the Darwinian process through which the best innovations rise above the rest and ultimately scale across the economy and society.

While there’s a commonly held belief that business investment is focused on the longer-term future and largely independent of current consumption, historical data shows this to be a myth. In virtually every US recession since the 1940s, investment has fallen precipitously.
³⁸
The investment decisions that businesses make are deeply influenced by both the current economic environment and the near-term outlook. In other words, tepid consumer demand today can rob us of prosperity in the future.

In an environment where consumers continue to struggle, many businesses will be inclined to focus on cutting costs rather than expanding markets. One of the few relative bright spots for potential investment is likely to be labor-saving technology. Venture capital
and research-and-development investment might then flow disproportionately into innovations specifically geared toward eliminating workers or deskilling jobs. At some point down the line, we could end up with plenty of job-seeking robots—but less of the broad-based innovation that improves the overall quality of our lives.

The trends we’ve examined in this chapter are all based on what I would characterize as a very realistic, and even conservative, view of the way technology is likely to progress. There can be little doubt that those occupations that primarily involve the execution of tasks that are relatively routine and predictable are going to be highly susceptible to further automation over the course of the next decade or so. As these technologies improve over time, more and more jobs will be impacted.

There is an even more extreme possibility, however. A great many technologists—some of whom are considered to be leaders in their fields—have a far more aggressive view of what will ultimately be possible. In the next chapter, we’ll take a balanced look at some of these truly advanced, and far more speculative, technologies. It may well be that these breakthroughs will remain science fiction for the foreseeable future—but if they are ultimately realized, that would dramatically amplify the risk of soaring technological unemployment and income inequality, and perhaps lead to scenarios even more dangerous than the economic risks we’ve focused on so far.

*
Not all robots are used in production, of course. There are also consumer robots. Suppose you someday own a personal robot, capable of doing things around the house. It may “consume” electricity and require repair and maintenance. However, in economic terms,
you
are the consumer—not the robot. You need a job/income or you won’t be able to pay for the operating costs of your robot. Robots don’t drive final consumption—people do. (Assuming, of course, that robots are not truly intelligent, sentient, and accorded the economic freedom that would be necessary for them to act as consumers. We’ll consider that speculative possibility in the next chapter.)

*
It’s important to note that retail sales are only a small fraction of overall consumption, or what is technically called personal consumption expenditure (PCE). PCE is usually around 70 percent of US GPD and includes all the products and services that consumers purchase, as well as housing expenditures—either rent or “imputed rent” (a measure used for owner-occupied dwellings).

*
Krugman’s primary objection relates to the fact that consumers at various points on the income distribution aren’t necessarily at that level all the time. Some people might be having an especially good or bad year, and their spending will be more a function of their long-term expectations than their current situation. (This, as we will see shortly, relates to what’s called the “permanent income hypothesis.”) As a result, Krugman says, looking at the data at any moment in time “tells you nothing at all about what will happen.” Krugman points out that “economics is not a morality play,” and goes so far as to suggest that we can have “full employment based on purchases of yachts, luxury cars, and the services of personal trainers and celebrity chefs.” I am skeptical of this (but see the section on “techno-feudalism” later in this chapter). As I pointed out previously, nearly all the major industries that constitute the modern economy produce mass-market products and services. Yachts and Ferraris just aren’t important enough to sustainably offset a broad-based reduction in demand for all the stuff that 99 percent of consumers buy. In any case, production of yachts and Ferraris will increasingly be automated. And how many personal trainers and celebrity chefs do the .01 percent really need?

*
This “fast food effect” may loom large for skilled workers in many other fields. Long before robots are able to completely replace these workers, technology may deskill the jobs and drive wages down. A classic example of deskilling involves London taxi drivers. Entering this profession requires memorizing an extraordinary amount of information about London’s street layout. This is referred to as “The Knowledge” and has been required of cab drivers since 1865. Neuroscientist Eleanor Maguire of University College London found that all this memorization actually resulted in changes to the drivers’ brains: London cabbies, on average, developed a larger memory center (or hippocampus) than people in other occupations. The advent of GPS-based satellite navigation has, of course, greatly reduced the value of all that knowledge. Taxi drivers possessing The Knowledge—who drive the famous “black” cabs (no longer black, but now covered in colorful advertising)—still dominate in London, but this is largely due to regulation. Drivers without The Knowledge have to be pre-booked; they are not allowed to be flagged down on the street. Of course, new services like Uber, which lets you book a cab with your smart phone, may soon make the act of flagging down a taxi itself obsolete. The taxi drivers may eventually be replaced completely by automated cars, but long before that happens, technology might well deskill their jobs and lower their wages. Perhaps regulation will save the London cabbies from this fate, but workers in many other fields will not be so lucky.

*
When a central bank like the Federal Reserve “prints money,” it normally purchases government bonds. When it settles the transaction, it deposits money into the bank account of whomever it bought the bonds from. This is newly created money: it just appears out of nowhere. Once this new money is in the banking system, the idea is that banks can then loan it out. This is what’s known as fractional reserve banking. Banks have to keep a small percentage of the new money on hand, but they’re allowed to loan out most of it. The way things are supposed to work is that the banks loan the new money to businesses that can then expand and hire more people. Or the banks might loan to consumers who spend the money, thereby creating new demand. Either way, jobs should be created and money (purchasing power) will flow to consumers. Eventually, the money once again gets deposited in a bank and then most of it can yet again be loaned out—and so on. In this way, the newly created money cascades through the economy, multiplying and generally being fruitful. However, if automation technology eventually makes it possible for businesses to expand or meet new demand without significant hiring, or if demand is so weak that businesses aren’t interested in borrowing, then little of the newly created money will find its way to consumers, and so it won’t get spent and it won’t multiply in the intended fashion. It will just slosh around in the banking system. This is more or less what occurred during the 2008 financial crisis—not because of job automation, but because the banks could not find creditworthy borrowers, and/or no one wanted to borrow anyway. Everyone just wanted to hold onto their cash. Economists call this situation a “liquidity trap.”