Rise of the Robots: Technology and the Threat of a Jobless Future (32 page)

In his 2001 book
Fast Food Nation,
Eric Schlosser relates how McDonald’s was already experimenting with more advanced labor-saving technology in the 1990s. At test sites in Colorado Springs, “robotic drink machines selected paper cups, filled them with ice, and then filled them with soda,” while French-fry cooking was fully automated and “advanced computer software essentially ran the kitchen.”
¹⁹
That all these innovations were not eventually scaled across McDonald’s restaurants everywhere may well have something to do with the fact that wages have remained very low. This situation cannot be expected to persist indefinitely, however. Eventually, technology will advance to the point where low wages no longer outweigh the benefits of further automation. Introducing more machines might also convey important benefits beyond simply reducing labor costs, such as improved quality or consistency or the consumer perception that automated preparation is more hygienic. As well, there might be synergies between robotic production and other emerging technologies. For example, today it’s easy to imagine a mobile app that allows customers to design a completely custom meal, pay for it in advance, and then expect it to be ready for pickup at a precise time; that would have been fantasy in the 1990s. The upshot of all this is that labor-saving technology in an industry like fast food is unlikely to advance in a consistent, predictable way. Instead, it may remain relatively stable for long periods and then leap forward rapidly once things reach a tipping point that forces a reevaluation of the worker-machine trade-off.

Still another consideration involves the behavior of consumers when they are faced with unemployment or reduced incomes. A change in income that consumers expect to be long-term or permanent will have a much bigger impact on their spending behavior than a short-term one. Economists have an impressive name for this idea—“the permanent income hypothesis”—and it was formalized by Nobel laureate Milton Friedman. For the most part, however, it amounts to simple common sense. If you win a thousand dollars in
the lottery, you might spend some of it and save the rest, but you’re unlikely to make a major, ongoing change to your spending behavior. After all, it’s only a one-time bump in your income. On the other hand, if you get a thousand-dollar raise per month, you might well lease a new car, start eating out more often, or even move to a more expensive home.

Historically, unemployment has been viewed as a short-term phenomenon. If you lose your job but feel confident of finding a new position at comparable pay within a short timeframe, you might choose to simply draw on your savings or use your credit card to continue spending at nearly the same level. During the postwar period, it was common for companies to lay off workers for a few weeks or months and then hire them back as soon as the outlook improved. The situation is obviously now quite different. In the wake of the 2008 financial crisis, the long-term unemployment rate soared to unprecedented levels, and it continues to be very high by historical standards. Even those experienced workers who manage to find a new job very often have to accept a lower-paying position. These realities are not lost on consumers. Accordingly, it seems reasonable to speculate that the perception of what it means to be unemployed may gradually be changing. As more people come to see unemployment as a longer-term—or in some cases perhaps even permanent—situation, this seems likely to amplify the impact of a job loss on their spending behavior. In other words, the historical record is not necessarily a good predictor of the future: as the implications of advancing technology become evident to consumers, they may choose to cut spending more aggressively than has been the case in the past.

The complexity that operates in the real-world economy is, in many ways, somewhat analogous to that of the climate system, which is likewise characterized by a nearly impenetrable web of interdependencies and feedback effects. Climate scientists tell us that, as the amount of carbon dioxide in the atmosphere increases, we should not expect a steady, consistent rise in temperatures. Instead, average
temperatures will advance chaotically in an uptrend punctuated by plateaus and, quite possibly, years or even longer periods that are relatively cool. We can also expect an increase in the number of storms and other extreme weather events. A somewhat similar phenomenon may unfold in the economy as income and wealth become progressively more concentrated and an ever larger fraction of consumers struggle with a dearth of purchasing power. Measures like productivity or the unemployment rate will not advance smoothly, and the likelihood of financial crises may well increase. Climate scientists also worry about tipping points. For example, one risk is that rising temperatures might cause the arctic tundra to melt, releasing huge amounts of sequestered carbon and, in turn, causing warming to accelerate. By a similar token, it’s possible that at some future point, rapid technological innovations might shift the expectations of consumers about the likelihood and duration of unemployment, causing them to aggressively cut their spending. If such an event occurred, it’s easy to see how that could precipitate a downward economic spiral that would impact even those workers whose jobs are not directly susceptible to the technological disruption.

Is Economic Growth Sustainable as Inequality Soars?

As we’ve seen, overall consumer spending in the United States has so far continued to grow even as it has become ever more concentrated, with the top 5 percent of households now responsible for nearly 40 percent of total consumption. The real question is whether that trend is likely to be sustainable in the coming years and decades, as information technology continues its relentless acceleration.

While the top 5 percent have relatively high incomes, the vast majority of these people are heavily dependent on jobs. Even within these top-tier households, income is concentrated to a staggering degree; the number of genuinely wealthy households—those that can survive and continue spending entirely on the basis of their
accumulated wealth—is far smaller. During the first year of recovery from the Great Recession, 95 percent of income growth went to just the top 1 percent.
²⁰

The top 5 percent is largely made up of professionals and knowledge workers with at least a college degree. As we saw in
Chapter 4
, however, many of these skilled occupations are squarely in the crosshairs as technology advances. Software automation may eliminate some jobs entirely. In other cases, the jobs may end up being deskilled, so that wages are driven down. Offshoring and the transition to big data–driven management approaches that often require fewer analysts and middle managers loom as other potential threats for many of these workers. In addition to directly impacting households that are already in the top tier, these same trends will also make it harder for younger workers to eventually move up into positions with comparable income and spending levels.

The bottom line is that the top 5 percent is poised to increasingly look like a microcosm of the entire job market: it is at risk of itself being hollowed out. As technology progresses, the number of American households with sufficient discretionary income and confidence in the future to engage in robust spending could well continue to contract. The risk is further increased by the fact that many of these top-tier households are probably more financially fragile than their incomes might suggest. These consumers tend to be concentrated in high-cost urban areas and, in many cases, probably do not feel especially wealthy. Large numbers of them have climbed into the top 5 percent through assortative mating: they have partnered with another high-earning college graduate. However, housing and education costs are often so high for these families that the loss of either job puts the household at substantial risk. In other words, in a two-income household the likelihood that sudden unemployment will lead to a substantial cut in spending is effectively doubled.

As the top tier comes under increasing pressure from technology, there are few reasons to expect that the prospects for the bottom
95 percent of households will improve significantly. Robotics and self-service technology in the service sector will continue to make inroads, holding down wages and leaving relatively unskilled workers with fewer options. Automated vehicles or construction-scale 3D printers may eventually destroy millions of jobs. Many of these workers may experience downward mobility; some will likely choose to leave the labor force entirely. There is a risk that, over time, more households will end up living on incomes that are very close to the subsistence level; we could well see even more shoppers in midnight lines waiting for their EBT cards to be reloaded so they can feed their families.

In the absence of increasing incomes, the only mechanism that will allow the bottom 95 percent to spend more would be to take on more debt. As Cynamon and Fazzari found, it was borrowing that allowed American consumers to continue driving economic growth over the course of the two decades leading up to the 2008 financial crisis. In the wake of that crisis, however, household balance sheets are weak and credit standards have tightened substantially, so a great many Americans cannot finance further consumer spending. Even if credit again begins to flow to these households, that is necessarily a temporary solution. Increased debt is unsustainable without increased income, and there would be an obvious danger that loan defaults might eventually precipitate a new crisis. In the one area where lower-income Americans still have easy access to credit—student loans—the debt burden has already grown to extraordinary proportions and the resulting payments will decimate the disposable income of college graduates (not to mention those who fail to get a degree) for decades to come.

While the argument I’m making here is theoretical, there is statistical evidence to support the contention that inequality can be harmful to economic growth. In an April 2011 report, economists Andrew G. Berg and Jonathan D. Ostry of the International Monetary Fund studied a variety of advanced and emerging economies and came
to the conclusion that income inequality is a vital factor affecting the sustainability of economic growth.
²¹
Berg and Ostry point out that economies rarely see steady growth that continues for decades. Instead, “periods of rapid growth are punctuated by collapses and sometimes stagnation—the hills, valleys, and plateaus of growth.” The thing that sets successful economies apart is the duration of the growth spells. The economists found that higher inequality was strongly correlated with shorter periods of economic growth. Indeed, a 10-percentage-point decrease in inequality was associated with growth spells that lasted 50 percent longer. Writing on the IMF’s blog, the economists warned that extreme income inequality in the United States has clear implications for the country’s future growth prospects: “Some dismiss inequality and focus instead on overall growth—arguing, in effect, that a rising tide lifts all boats.” However, “when a handful of yachts become ocean liners while the rest remain lowly canoes, something is seriously amiss.”
²²

Long-Term Risks: Squeezed Consumers, Deflation, Economic Crises, and . . . Maybe Even Techno-Feudalism

After I published my first book on the subject of automation in 2009, several readers wrote to me to point out that I had neglected to focus on an important point: robots might indeed drive down wages or cause unemployment, but more efficient production would also make everything much cheaper. So even if your income fell, you’d still be able to continue consuming since prices for the things you wanted to buy would be lower. This seems to make sense, but there are a few notable caveats.

The most obvious issue is that many people might be unemployed entirely and effectively have zero income. In that situation, low prices don’t solve their problem. Additionally, some of the most important components of the average household budget are relatively immune to the impact of technology, at least in the short and medium terms. The
cost of land, housing, and insurance, for example, are tied to general asset values, which are in turn dependent on the overall standard of living. This is the reason that developing countries like Thailand don’t allow foreigners to buy land; doing so might result in prices being bid up to the point where housing would become unaffordable for the country’s citizens. As we saw in
Chapter 6
, health care costs also probably represent a challenge for the robots in the near term. Automation is likely to have the greatest immediate impact on costs in manufacturing and in some discretionary services, especially information and entertainment. Yet, these things are a relatively small part of most household budgets. The big-ticket items—housing, food, energy, health care, transportation, insurance—are much less likely to see rapid, near-term cost reductions. There’s a real danger that households will end up being squeezed between stagnant or falling incomes and major-expense items that continue to rise in cost.

Even if technology does eventually manage to reduce prices across the board, there is a critical problem with this scenario. The historical path to prosperity has generally been one of wages increasing faster than prices. If someone from the year 1900 were to travel forward in time and visit a contemporary supermarket, he or she would, of course, be shocked by the high prices. Nonetheless, we now spend a significantly smaller share of our incomes on food than was the case in 1900. Food has become cheaper in real terms even as nominal prices have increased dramatically. This has happened because incomes have increased even more dramatically.

Now imagine the opposite situation: incomes are falling, but prices are falling even faster. In theory, this would also mean your purchasing power was increasing: you should now be able to buy more stuff. In reality, however, deflation is a very ugly economic scenario. The first problem is that a deflationary cycle is quite hard to break. If you know that prices will be lower in the future, why buy now? Consumers hold back, waiting for even lower prices, and that in turn forces even more price cuts as well as reduced production of
goods and services. Another problem is that, in practice, it’s often difficult for employers to actually lower wages. Instead, they are more likely to cut workers, so deflation is typically associated with soaring unemployment, and again, that eventually leads to a lot of consumers with no income at all.