The Internet is one of humanity’s greatest technical advances. Yet compared to great technological inventions of the past, it is also a colossal economic disappointment.
I’m talking about jobs.
Yes, young programmers are getting jobs straight out of college at salaries in the six figures. But I’m referring to jobs in a deep and sustaining sense — employment well beyond the “1 percent.”
For all its economic virtues, the Internet has been long on job displacement and short on job creation. As a result, it is playing a central role in wage stagnation and the decline of the middle class.
Sure, the Internet has created new applications and great companies — Google, Facebook, Amazon, Twitter, and the all-important cloud. But many of the largest Internet companies have for the most part taken revenue from existing companies without growing the total economy.
The technologies of the past had massive new job creation effects that swamped displacement effects. The Internet on the other hand has massive displacement effects that are overwhelming the job creation effects. In the past, new technological achievements created new industries that not only absorbed the displaced workers but generated opportunities for many more. The result was a vibrant middle class.
Consider the integrated circuit, which first appeared on the market in 1961. At that time, the worldwide electronics market was $29 billion. Today it is on the order of $1.5 trillion. The integrated circuit made existing products better. For example, vacuum tube mainframe computers were replaced by computers based on integrated circuits. The new machines were less expensive, far faster, more reliable, substantially smaller, and much more energy efficient. As a result the mainframe computer business expanded rapidly. IBM’s revenue increased from less than $2 billion in 1960 to over $26 billion in 1980. The integrated circuit also spawned new industries and applications that never existed before — cellular communications, PCs, tablets, and the Internet of Things.
The story of the internal combustion engine is even more dramatic. Not only did it create the automotive industry, but Henry Ford shocked the industrial world when he doubled the pay of assembly line workers to $5 a day. Ford reasoned that a higher paid workforce would be able to buy more cars and thus would grow his business. Others followed suit. Ford’s action helped to create the middle class.
Automotive companies also created a large demand for other products and services that employed millions more — steel, coal to make the steel, glass, machine tools, auto dealers and dealerships, gas stations, oil fields, mechanics, bridges, roads, construction equipment, etc. Automobiles created suburbia and the home construction boom that followed. They made a new form of retail distribution possible—the shopping center. The
in new jobs purchased homes, appliances, and clothes creating still more jobs. During the 20th Century, the industrialized world enjoyed the fruits of what economists call the virtuous circle.
To date the Internet has been much more effective at eliminating jobs than creating new ones. Exhibit A: Online retailing has directly replaced many jobs and indirectly eliminated many more. Amazon’s extremely efficient distribution system replaces retail stores and their employees. Their warehouses use robots instead of workers.
Those are the direct effects. The indirect effects are the disappearing need for retail space, along with workers who build the stores and maintain them, as well as companies that supply retail establishments with furnishings.
The Internet has made shopping more efficient and created more competition that has driven down consumer prices. But it has had little or no effect on per capita sales. Monthly retail sales adjusted for both inflation and population growth are below where they were prior to the 2008 recession — $165 versus $168 billion — and have increased by less than 10% in the last 15 years or about 0.6% per year. Meanwhile, employment in the retail and wholesale trade has dropped from about 21.2 million in 2000 to 19.9 million in 2010.
Those highly paid young coders are a select few. They are also a symptom of something more insidious: The Internet is so efficient that it can create large income companies with few employees.
The reason Google, Facebook, and Twitter can pay them such large salaries is that the Internet companies is so efficient they can generate high revenues with few employees.
In 2013, Google had around 50,000 employees and generated revenues of around $55 billion in sales or about $1.0 million per employee. The numbers are similar for Facebook. Amazon was running at an $74 billion revenue rate and had around 110,000 employees or a little over $670,000 in sales per employee.
In the United States, each non-farm worker adds a little over $120,000 to the domestic output. That means that highly productive Internet companies must create five to ten times the dollars in sales to justify hiring an employee as the average company of the past did.
The prevailing economic wisdom is that new technologies will create new opportunities that will offset the effects of displacement. We continually use the experiences of the past to support our hopes about the future. But the experiences of the past took place in the physical world. Our future will be increasingly played out in a virtual one.
Given that the Internet isn’t turning out to be the job creation engine of the future we all hoped for, we had better get to work on searching for and implementing policies that will offset the Internet’s displacement effects.
To start with those policies must be implemented with the Internet’s efficiency in mind. Raising the minimum wage, for instance, plays straight into the hands of the Internet efficiency engine. Raising the minimum wage will just drive employers to use machines to replace people. An earned income tax credit is a better approach. Low paid workers get the benefit of transfer payments and employers who will not pay hirer wages will feel less pressure to automate.
Investing in infrastructure is an excellent way to create jobs but such infrastructure should be compatible with an increasingly virtual world. Yes we should fix the roads but as more and more people work from home, as more and more of what we purchase gets delivered to our doorstep, as more and more of us go out to the movies in our living rooms, and as highway congestion grows, the chances are that more and more of us will use our cars less.
Millennials are the harbingers of this new trend. The numbers of cars purchased by people 18 to 34 years old has fallen by almost 30%. Millennials are opting to spend their money on high tech things like tablets, smart phones, and high bandwidth access.
For a millennial, the infrastructure of the future will be higher bandwidth interconnections and public transportation that will take the place of his car.
Actions like these will chip away at the problem. The challenge will be to find enough to them to offset the effects of the most powerful efficiency engine the world has ever known.