Introduction

Growing inequality is a hot topic. There has recently been an explosion of new books on the subject. The following have been published in 2014 alone:

• “Capital in the Twenty-First Century” by Thomas Piketty and Arthur Goldhammer;
• “The Price of Inequality: How Today's Divided Society Endangers Our Future” by Joseph Stiglitz;
• “Divided: The Perils of Our Growing Inequality” by David Cay Johnston;
• “The Divide: American Injustice in the Age of the Wealth Gap” by Matt Taibbi and Molly Crabapple;
• “Degrees of Inequality: How the Politics of Higher Education Sabotaged the American Dream” by Suzanne Mettler.

Many of these books draw on history for explanations of inequality with citations of Simon Kuznets - “Toward A Theory Of Economic Growth”, Karl Marx - “Das Kapital” and  Joseph Spengler – “Decline of the West”. Others view the widening gap as a reflection of the growing influence and power of the “moneyed” classes. And some talk of conspiracy.

In my view, these are all secondary. The single primary reason for the growing inequality is “labor saving” resulting from the Information Revolution. New information technologies are replacing large blocs of middle class jobs and with more people out of work, wages have plummeted. Weaker unions? Maybe, but there is very little unions can do when the demand for their members falls as much as it has.

What has already happened coupled with future applications of information technologies will make an even larger segment of middle class workers redundant. And this raises a larger question: if the “work” of a large portion of the middle class is not needed, where does their income come from? This is a serious problem because the global societal norm is that people earn income by working. In what follows, I offer documentation of my claim and discuss the broader question.

The Information Revolution

Remember the 1995-2000 “dot-com bubble”? It was based on the belief that the new information technologies had arrived and were going to revolutionize how we lived and worked. It failed. Nobody questioned the potential of the new technologies, but there were two problems:

We had not yet really figured out how to use the new technologies effectively, and the history of science tells us it takes humans a while to adapt to new technologies. That was back then. It is now a decade later, and the new technologies are now being effectively utilized in all aspects of life and work. And remember too that recessions/recoveries give employers a great chance to rethink how they operate and whether they need to hire all their workers back. What in essence are the new technologies providing? Two things:

• the ability to perform manual labor better and more efficiently than humans, e.g., factory work, and
• the ability to process information better and more efficiently than humans, e.g., bill pay.

In short, the new technologies are very labor-saving.

The Big Picture

As can be seen (Table 1), US employment in durables goods production not grown much since 1947. But over this same period, the US Bureau of Economic Analysis estimates that durable goods value added in constant prices has increased ten times. That means a worker today can produce ten times more than what a worker could produce in 1947.

Table 1. – US Employment Changes, 1947 - 2013

Source: US Bureau of Labor Statistics

What Has Happened in Manufacturing

We have heard a lot about US jobs being lost to overseas producers: for example, a quote from an earlier piece:

“One of the biggest challenges facing the American economy is that we lack a domestic manufacturing base. Simply put we do not produce anything anymore. We import most of our goods which has resulted in a huge trade deficit and industrial job losses.”

Nonsense.

Productivity gains and not overseas outsourcing is the primary reason for declining jobs in manufacturing. In a recent study, The McKinsey Global Institute came to the same conclusion:

“…overall in the United States, trade and outsourcing explain only about 20 percent of the 5.8 million manufacturing jobs lost during the 2000-10period; more than two-thirds of job losses can be attributed to continued productivity growth, which has been outpacing demand growth for the past decade.”

What sort of productivity gains are we talking about? Data from the US Bureau of Labor Statistics indicate that output per labor hour in manufacturing has grown at a compounded rate of 4.79% since 1947. Reliable data on making autos does not go back that far, but suppose it did. 4.79% compounded annually would mean that if 10 workers could assemble 10 cars per day in 1947, they could assemble 78 autos today!

Pictures are worth thousands of words. Look at following pictures: the “crowds” of workers at River Rouge versus the “loneliness” of the BMW plant. How could unions or any other worker group fight what was happening.

Ford River Rouge Plant – Circa 1939

BMW Plant - 2014

US Agriculture: Labor Saving Has Already Happened

According to US Department of Agriculture data, US agricultural output increased at a 1.49% average annual rate between 1948 and 2011. Over that same period, agricultural employment fell at a 2.41% average annual rate. The 2012 Census of Agriculture reports there are now only 3.2 million farm workers managing 914.5 million acres. Similar productivity gains can be expected from in other sectors from future applications of the information revolution.

Where Is All This Going?

Table 1 showed that employment in services has grown at a 2.29% compounded annual rate since 1947. And while experience suggests that guesses on where the evolving information revolution will take us are dangerous, it is hardly a stretch to say this growth will not continue. Information technology innovations in services are starting to take their toll. Table 2 shows the leading employers today by industry.

Table 2. – US Employment, by Industry, 2013 (in thousands)

Source: US Bureau of Labor Statistics

Consider first sales (retail and wholesale trade) and the impact of online business. Retailers and wholesalers are increasingly making sales via the Internet. Cases in point:

• Retailers for most clothing, shoes, office supplies, books, hardware supplies and movies are losing markets to online suppliers. The online sales share will continue to grow: online sellers can maintain much larger inventories at costs than retailers. The consequence will be job losses in retail.
• Drug stores: In addition to providing prescriptions, drug stores have become large convenience stores. Most goods are branded. Aside from cosmetics, there is really nothing they sell that could not be sold online. Certainly having doctors send prescriptions to drug companies for online fulfillment would be much more efficient and preferable to current practices. Retail job losses can be expected here as well.
• Consider next real estate, investments, and autos. Ten years ago, realtors, stock brokers and auto salesmen provided buyers with nearly all the information on these purchases: they were needed. Today, this information is online. Local realtors, brokers, and auto sales people will lose a lot of business going forward to online sellers.
• Gambling: at least in the US, being able to gamble online has reduced attendance at gambling locales such as horse races and casinos.
• Sporting events: nearly all these events are delivered to consumers online. And how much longer will these events be played before sell-out crowds? How many more terrorist bomb events have to occur (Boston Marathon) before people start staying away from such events?
• Gas stations – it is not convenient to buy gas online. But switching to electric plug-in autos would change this.

There are some items shoppers will want to see physically before buying. These include fresh food, goods with an aesthetic dimension and unbranded items. While strip malls have an uncertain future, visits to indoor malls, especially in the winter, have become a major form of recreation for American families.

In health, information technology advancements have already had a tremendous impact: the ability to transfer and analyze health records will lead to major improvements in health service deliveries. And certainly the Internet allows people to be better-informed about their problems. However, health services have a significant personal component and consequently, significant job losses are not likely.

In education, the online element is growing. Significant labor-saving will result if online lectures by outstanding speakers are combined with follow-up teacher and student discussions.

Conclusion and What to Do?

In sum, labor saving technologies resulting from the information revolution have wiped out a huge segment of middle class jobs. This is the primary reason for growing income inequality in the US.  

In earlier times, people have worried that advances in labor-saving technology would eliminate too many jobs. Economists always scoffed and said wages will adjust downward until full employment was reached. And that always happened. But now, as an economist, I am not so sure. And yes, there might be a price where unneeded workers would be employable. But that “price” (wage) might be so low as to make income inequality unacceptably large.

A little history is in order. Since the beginning of time, compensation has been based on trading things “of value” For example, my work for the food you grow. But what if “work” of a large segment has no value? A real problem.

Juliet Schor, an economist and sociologist has been thinking and writing about these issues for some time. She points out that historically, civilizations work fewer hours as they develop. Information on a number of developed countries is provided in the following graph. Even over this period, one can see a decline but with an uptick at the end of the period as people work more hours after capital losses in the global recession.

Average Hours Worked per Year, Selected Countries

Source: OECD

In light of the latest spate of labor-saving technologies, is this gradual decline hours worked enough? Should the US be doing more to get people to work less? Schor makes the following important point about current policies to reduce unemployment:

“Trying to grow our way out of technical-change induced unemployment is ecological suicide. We’re already producing more pollution than the climate, and planet, can tolerate. This will have us spending money on things we don’t need."

Schor recommends doing things that open up the labor market. We could expand, rather than contract, social security eligibility. Smart countries with unemployment problems try to get senior workers to retire earlier. They don’t create incentives for them to work longer. We could give partial benefits for people to gradually reduce work. Schor points out that shorter hours of work are possible at many levels–more schooling at the beginning of the work life, four day workweeks, and then tapering off hours at the end of the work life.

We are facing a real problem. There are no easy answers.