Rise of the Robots: Technology and the Threat of a Jobless Future (31 page)

In the wake of the Great Recession, Walmart stores routinely see an explosion of activity just after midnight on the first of each month—the day that electronic benefits transfer (EBT) cards are reloaded by the government. By the end of the month, Walmart’s lowest-income customers have quite literally run out of food and other essentials, so they load up their shopping carts and line up in anticipation of a credit from the food stamp program that generally comes through shortly after midnight.
⁹
Walmart has also suffered from increased competition from dollar stores; in many cases its customers are turning to these outlets not necessarily because overall prices are lower but, rather, because the stores offer smaller
quantities that help them stretch their few remaining dollars as they struggle to make it through the final days of the month.

Indeed, throughout the private sector, the recovery has largely been characterized by soaring corporate earnings coupled with often underwhelming revenues. Corporations have achieved dizzying levels of profitability, but they have accomplished this primarily by cutting labor costs—not by selling more of the goods and services they produce. This shouldn’t come as a surprise: take a moment to look back at
Figures 2.3
and
2.4
in
Chapter 2
. Corporate profits as a share of GDP reached unprecedented heights even as labor’s share of national income plunged to a record low. To me, this suggests that a great many American consumers are struggling to purchase the products and services that companies are producing.
Figure 8.1
, which shows how general US corporate earnings recovered rapidly and have been pulling away from retail sales over the course of the recovery, makes the story still more clear.
*
Keep in mind that, as we saw previously, the gradual spending recovery has been powered entirely by consumers in the top 5 percent of the income distribution.

The Wisdom of the Economists

Despite the evidence suggesting that a huge percentage of American consumers simply don’t have sufficient income to create adequate demand for the products and services produced by the economy, there is no general agreement among economists that income inequality is creating a substantial drag on economic growth. Even among America’s leading progressive economists—nearly all of whom would likely agree that a lack of demand is a primary problem
facing the economy—there is no consensus about the direct impact of inequality.

Figure 8.1. US Corporate Profits Versus Retail Sales During Recovery from the Great Recession

S
OURCE
: Federal Reserve Bank of St. Louis (FRED).
¹⁰

The Nobel laureate economist Joseph Stiglitz has been perhaps the most vocal proponent of the idea that inequality undermines economic growth, writing in a January 2013
New York Times
op-ed that “inequality is squelching our recovery” because “our middle class is too weak to support the consumer spending that has historically driven our economic growth.”
¹¹
Robert Solow—who won the Nobel Prize in 1987 for his work on the importance of technological innovation to long-term economic growth—seems to largely agree, saying in a January 2014 interview that “increasing inequality tends to hollow out the income distribution, and we lose the solid middle class jobs and steady middle class incomes which provide a reliable flow of consumer demand that keeps industry going and innovating.”
¹²
Paul Krugman, yet another Nobel laureate—and the one with the highest profile as a columnist and blogger for the
New York Times
—disagrees, however, writing in
his blog that he wishes he “could sign on to this thesis,” but that the evidence doesn’t support it.
¹³
*

Among more conservative economists, the idea that inequality is a significant drag on growth is likely to be dismissed entirely. Indeed, many right-leaning economists are reluctant even to accept the argument that a lack of demand has been the primary problem facing the economy. Instead, throughout the course of the recovery, they have pointed to uncertainty surrounding issues like public debt levels, potential tax increases, increased regulation, or the implementation of the Affordable Care Act. Cutting government spending and reducing taxes and regulation, they say, will spur investor and business confidence, leading to increased investment, economic growth, and employment. This idea—which seems to me to be remarkably divorced from the obvious reality—has been repeatedly disparaged by Krugman as a belief in “the confidence fairy.”
¹⁴

My key point here is that professional economists—all of whom have access to the same objective data—are completely unable to agree on what I would characterize as an extraordinarily
fundamental economic question: Is a demand shortfall holding back economic growth, and if so, is income inequality an important contributor to the problem? I suspect that the lack of consensus on this question offers a pretty good preview of what we can expect from the economics profession as the technological disruption I’ve been describing in these pages unfolds. While it’s certainly possible that two “scientists” may look at the same data and interpret it differently, in the field of economics the opinions all too often break cleanly along predefined political lines. Knowing the ideological predisposition of a particular economist is often a better predictor of what that individual is likely to say than anything contained in the data under examination. In other words, if you’re waiting for the economists to deliver some sort of definitive verdict on the impact that advancing technology is having on the economy, you may have a very long wait.

Beyond the ideological divide in economics, yet another potential problem is the extreme quantification of the field. In the decades since World War II, economics has become extraordinarily mathematical and data-driven. While this certainly has many positive aspects, it is important to keep in mind that there is obviously no economic data streaming in from the future. Any quantitative, data-driven analysis necessarily depends entirely on information gathered in the past, and in some cases, that data may have been collected years or even decades ago. Economists have used all that past data to construct elaborate mathematical models, but most of these trace their origin to the economy of the twentieth century. The limitations of the economists’ models were made evident by the near-total failure of the profession to anticipate the 2008 global financial crisis. In a 2009 article entitled “How Did Economists Get It So Wrong?” Paul Krugman wrote that “this predictive failure was the least of the field’s problems. More important was the profession’s blindness to the very possibility of catastrophic failures in a market economy.”
¹⁵

I think there are good reasons to be concerned about a similar failure of the economists’ mathematical models as the exponential
advance of information technology increasingly disrupts the economy. Adding to the problem is that many of these models employ simplistic—and in some cases seemingly absurd—assumptions about the way consumers, workers, and businesses behave and interact. John Maynard Keynes may have said it best, writing nearly eighty years ago in
The General Theory of Employment, Interest and Money,
the book that arguably founded economics as a modern field of study: “Too large a proportion of recent ‘mathematical’ economics are merely concoctions, as imprecise as the initial assumptions they rest on, which allow the author to lose sight of the complexities and interdependencies of the real world in a maze of pretentious and unhelpful symbols.”
¹⁶

Complexity, Feedback Effects, Consumer Behavior, and “Where Is That Soaring Productivity?”

The economy is an enormously complex system, ripe with a myriad of interdependencies and feedback loops. Change one variable and a variety of effects are likely to cascade through the system, some of which may act to mitigate or counteract the initial change.

Indeed, this propensity for the economy to self-moderate through feedback effects is likely one important reason that the role advancing technology has played in creating inequality remains subject to debate. Economists who are skeptical about the impact of technology and automation often point to the fact that the rise of the robots is not obvious in the productivity data, especially over the short term. For example, in the final quarter of 2013, productivity in the United States fell to an annualized rate of just 1.8 percent, as compared to a much more impressive 3.5 percent in the third quarter.
¹⁷
Recall that productivity is measured by dividing the economy’s output by the number of hours worked. So if machines and software were indeed substituting for human labor at a rapid clip, you would expect the number of hours worked to fall precipitously—and productivity, in turn, to soar.

The problem with this assumption is that in the real economy, things are not so simple. Productivity does not measure how much a business
could
produce per hour; it measures how much a business actually does produce. In other words, productivity is directly influenced by demand. Output, after all, makes up the numerator of the productivity formula. This is especially important when you consider that most of the economy in developed countries is now made up of service businesses. While a manufacturing company, faced with slack demand, might conceivably choose to keep cranking out products and letting them pile up in inventory or in distribution channels, a service business cannot do this. Within the service sector, output responds immediately to demand, and any business that experiences weak growth in demand for its output is likely to also experience less than impressive productivity growth, unless it immediately cuts its workforce or reduces worker hours sufficiently to keep the numbers in line.

Imagine you own a small business that provides some type of analytic service to large corporations. You have ten employees who are fully engaged. Suddenly, a powerful new software application appears that will allow just eight workers to do the work formerly performed by ten. So you purchase the new software and eliminate two jobs. The robot revolution is at hand! Productivity is poised to soar. But, wait. Now your most important client forecasts a downturn in demand for its own product or service. The contract you were supposed to sign this week never materializes. The near-term future looks grim. You just had a layoff, so you don’t want to demoralize your workforce by immediately cutting still more jobs. Before you know it, your eight remaining employees are spending a big chunk of their time watching YouTube videos on your dime. Productivity is tanking!

In fact, this was normally what happened during most past downturns in the United States. Recessions typically saw declining productivity because output fell more than hours worked. However, during the Great Recession of 2007–2009, the opposite happened:
productivity actually increased. Output fell substantially, but hours worked fell even more as businesses very aggressively slashed their workforces, increasing the burden on the remaining workers. The workers who kept their jobs (who certainly feared more cuts in the future) probably worked harder and reduced any time they spent on activities not directly related to their work; the result was an increase in productivity.

In the real economy, of course, scenarios like this play out in countless organizations of all sizes. Somewhere, a firm may be incorporating new technology that increases productivity. Elsewhere another firm may be cutting output in response to slack demand. Averaged together, they result in only a middling overall productivity number. The point is that short-term economic numbers like productivity are likely to be variable and somewhat chaotic. Over the long run, however, the trend will be far more clear. Indeed, we saw evidence for this in
Chapter 2
; recall that productivity has significantly outpaced wages since the early 1970s.

The impact of weak consumer demand on productivity is just one example of the kind of feedback effect that operates in the economy. There are many others, and they can act in both directions. For example, less than robust consumer demand can also slow the development and adoption of new technology. When businesses make investment decisions, they factor in both the current and the anticipated economic environment. When the outlook is poor or when profits decline, investment in research and development or in new capital expenditures is also likely to fall. The result is that technological progress in subsequent years may be slower than it otherwise would have been.

Another example involves the relationship between labor-saving technology and the wages of relatively unskilled workers. If advancing technology (or some other factor) causes wages to stagnate or even fall, then from management’s perspective labor will—at least for a time—become more attractive relative to machines. Consider
the fast food industry. In
Chapter 1
, I speculated that this sector may soon be ripe for disruption as advanced robotic technology is introduced. But this suggests a basic question: Why hasn’t the industry already incorporated more automation? After all, putting together hamburgers and tacos hardly seems to be on the forefront of precision manufacturing. The answer, at least in part, is that technology has indeed already had a dramatic impact. While machines have not yet completely substituted for fast food workers on a large scale, technology has deskilled the jobs and made the workers largely interchangeable. Fast food workers are integrated into a mechanized assembly-line process with little training required.
*
This is why the industry is able to tolerate high turnover rates and workers with minimal skill levels. The effect has been to keep these jobs firmly anchored in the minimum-wage category. And in the United States, after adjusting for inflation, the minimum wage has actually fallen more than 12 percent since the late 1960s.
¹⁸