Is the Internet Increasing or Decreasing Productivity?
As debate over productivity rages furiously, it is time to adopt better measures to capture the value technology adds to the economy.
It is an open secret that technological advances of the computer age have changed our lives in many ways. However, we have yet to answer a simple question: How much value have these advances created?
There are two opposing views. In a tour de force, Ezra Klein of Vox has examined both arguments brilliantly. On one side are internet bears who point out that productivity has been falling and the internet is not as big a deal as it is made out to be. On the other side are internet bulls who argue that the internet has unleashed a new wave of productivity that is not captured by current measures accurately.
The bears are led by Professor Robert Gordon, the celebrated author of The Rise and Fall of American Growth. In his magisterial work, Gordon describes how newborn children in 1820 “entered a world that was almost medieval.” In this “dim world lit by candlelight,” folk remedies treated heart problems and the fastest anyone could travel was “by hoof or sail.”
As per Gordon, the great-grandchildren of the 1820 generation inherited a world transformed. They could “create light with the flick of a switch instead of the strike of a match.” Buildings could extend vertically, not just horizontally. Metal ships, trains, motor vehicles, washing machines and modern ovens changed life as we know it. Machines replaced human labor and “no longer did society have to allocate a quarter of its agricultural land to support the feeding of the horses or maintain a sizable labor force for removing their waste.”
This golden age from 1870 to 1970 saw technological progress on steroids. Medical advancements such as sanitation, anesthetic, antibiotics, surgery, chemotherapy and antidepressants transformed life. Vaccination worked miracles. Killers like smallpox vanished. Both life expectancy and the global population grew dramatically.
Since then, the pace of technological progress has slowed. Gordon points out that, after 1970, growth in total factor productivity (TFP), the metric that captures innovation, has only been 0.57% a year, less than a third of the 1.89% rate of 1920-70. During this period, the TFP was abnormally high because of the second industrial revolution “when fossil fuels, the internal-combustion engine, advanced metals and factory automation came together to produce electric lighting, indoor plumbing, home appliances, motor vehicles, air travel, air conditioning, television and much longer life expectancy.” Apart from TFP, gross domestic product (GDP), output per person and output per hour also increased in these five decades.
For Gordon, the third industrial revolution of information and communication technology that we are currently experiencing is not as significant as the second one. The internet has “had little effect on purchases of food, clothing, cars, furniture, fuel and appliances.” We do not eat computers, wear them or drive them to work. In Gordon’s words, we still “live in dwellings that have appliances much like those of the 1950s and we drive in motor vehicles that perform the same functions as in the 1950s, albeit with more convenience and safety.”
He further argues that America’s “productivity growth will be further held back by the headwinds of rising inequality, stagnating education, an aging population, and the rising debt of college students and the federal government.” It is a grim picture and even some in Silicon Valley agree.
The most prominent among internet bears is Peter Thiel, the first investor in Facebook and the godfather of the PayPal mafia. He believes the definition of technology has narrowed to shiny amusements on the internet and has famously proclaimed, “We were promised flying cars; we got 140 characters.” In his view, housing in San Francisco is 50-60 years old, the subways in New York are over 100 years old, and even the planes we fly on are little different from 40 years ago. The laptop and smartphone screens we have become addicted to now distract us from our very surroundings.
THE IDEA OF CONSUMER SURPLUS
Naturally, the bulls do not buy into Gordon’s argument though. Marc Andreessen, a legendary venture capitalist and the creator of the first internet browser, makes the case that productivity might be higher than it seems. In his view, “rapid tech might not show up in GDP or productivity stats.” Hal Varian, the chief economist at Google, wonders whether productivity figures are measuring the right things. So does Bill Gates, who told Klein that productivity figures are not good at capturing service-type improvements such as “digital photos, easier hotel booking, cheap GPS and nearly costless communication with friends.”
Klein has sympathy for this argument. He points out that measures of productivity are based on the sum total of goods and services in an economy for sale. Google Maps has changed Klein’s life because he has “a crap sense of direction” and he would pay hundreds of dollars for it. However, he pays nothing and thus GDP data might be undercounting the value of Google Maps. As per Varian, this is a systematic problem in measuring GDP: “It’s good at catching value to businesses but bad at catching value to individuals.” This gap between what Klein pays for Google Maps and what he gets from it is what economists call “consumer surplus.”
Silicon Valley constantly trumpets this idea of consumer surplus. Far too many of the great new technologies such as Skype, Google Calendar or Dropbox are free or nearly free for the consumer. People bought Henry Ford’s cars and their purchases showed up in GDP figures. They do not buy the various products on the internet today, which therefore do not show up in these figures.
Ilya Talman, Xiao Qiao, Steve Rubinow, Michael Lippitz and Peter Evangelakis make this argument in a 2017 paper titled “The New Macroeconomics of Consumer Surplus.” They point out that the standard way to assess economic growth is through measuring the GDP. However, prices for digital products and services have been falling. In Andreessen’s words, they are “deflationary in nature.” As Klein argues, they push prices down, not up, confounding GDP measurements.
Talman and his co-authors agree with this argument. They point out that businesses are paying, say, $1,000 for a computer today instead of $10,000 they would have been willing to pay 10 years ago. Prices of personal computers have fallen even as their processing power and storage capacity have increased. To be fair, US statistical agencies use hedonic methods “for products whose characteristics and/or quality are changing rapidly (e.g. ICT goods).” Yet Talman et al. believe this method greatly underestimates the impact of exponential technology improvements. They calculate that the computing-cost adjusted real GDP growth from 2006 to 2015 would be 6% per year instead of the reported 2.8% per year.
The authors argue that measuring consumer surplus is difficult because consumers do not tell anybody what they would have been willing to pay for the product or service they use for free or for cheap. However, studies to capture consumer surplus have been published recently. As per Peter Cohen, Robert Hahn, Jonathan Hall, Steven Levitt and Robert Metcalfe, “the UberX service generated about $2.9 billion in consumer surplus” in four American cities in 2015. They calculate “the overall consumer surplus generated by the UberX service in the United States in 2015 [to be] $6.8 billion.”
Talman et al. quote a paper by Eric Brynjolfsson and Joo Hee Oh from the Massachusetts Institute of Technology (MIT) to prove their point. The MIT scholars argue that “there has been an explosion of digital services on the Internet, from Google and Wikipedia to Facebook and YouTube,” but “the value of these innovations is difficult to quantify, because consumers pay nothing to use them.” They “estimate the increase in consumer surplus created by free internet services to be over $100 billion per year in the US alone.”
The paper by Talman et al. goes further than other papers on the subject and posits a new way to calculate total consumer surplus. It suggests that “a given product market during a multi-year period is comprised of three parts: the consumer surplus in the year before the period begins (i.e., the baseline year), the annual changes in consumer surplus attributable to changes in willingness to pay, and the annual changes in consumer surplus attributable to changes in price.”
The authors compare the current consumer surplus in the digital economy to the one in the electricity market from 1901 to 1956. Prices in electricity constantly fell during this period, creating an unrecorded consumer surplus. Today, they argue that falling prices of iPhones are creating a similar consumer surplus. In fact, they estimate this figure to be much higher than $200-$300 billion per year that Professor Chad Syverson came up with in a paper in 2016.
Like Andreessen, Varian and Gates, Talman et al. take the view that “technological progress is driving important changes to the economy and society.” From iPhones to artificial intelligence to gene sequencing, these changes are as profound as the changes that occurred before. The authors argue for creating a new reliable measure of the value of technology that captures these changes better by supplementing key measures such as GDP figures with consumer surplus. Echoing Andreessen and other internet bulls, they argue that even low-income Americans today live better than kings and queens just a couple centuries ago because of technological progress. Talman and his co-authors believe this progress may also explain “why measured inflation in developed countries has remained very low despite unprecedented money-printing by central banks since 2008.”
CONSUMER DEFICIT AND NEGATIVE EXTERNALITIES
Internet bulls definitely have a point when they argue that value of technology might be underestimated. However, they miss out on the idea of a consumer deficit, which simply put is being forced to pay more for something than it was worth to the consumer. For instance, driving in the San Francisco Bay Area costs a fair bit. Too much gas is wasted stuck in traffic jams on Highway 101. In many other metropolitan areas of the world, people pay less for public transport that gets them from one point to another quicker and cheaper. Google Maps might have made driving more convenient in the Bay Area, but it has not solved the consumer deficit problem of paying more for getting around both in money and time.
Apart from consumer deficit, there are other losses. When people take public transport, they can read, answer emails or simply nod off. Driving in traffic jams takes a toll on people’s energy. It is bad for their physical and mental health. The time spent driving is a net dead weight loss and must have a negative impact on productivity that is not captured in figures of consumer surplus or deficit.
There is another big question that technology bulls fail to factor in: of what economists call externalities. If an activity affects other parties without being reflected in the cost of the goods and services, it is an externality. Sometimes externalities can be positive. For example, a private garden near your home might increase its value. At other times, externalities can be negative. A pig sty near your home is likely to lower its value.
In the case of the internet, many externalities are negative. Employees are often distracted during work and spend time amusing themselves on their laptops or phones. Students often waste copious amounts of time on Snapchat, dating apps and pornographic sites instead of doing their work. Measuring consumer surplus fails to account for the loss that employees and students are causing to their employers, parents or the economy, both in the short run and the long run. We have yet to come up with accurate methods to measure the fall in their productivity.
Furthermore, it is an open secret that social media is distracting and smartphones are addictive. A 2017 paper published by the National Institute of Health found that “habitual involvement with these devices may have a negative and lasting impact on users’ ability to think, remember, pay attention, and regulate emotion.” Most adults confess that they no longer read as they used to. Teens are in a worse condition and are facing a mental health crisis. These surely cause a fall in productivity and do not show up in GDP or consumer surplus figures adequately.
Economists like Syverson caution against the thesis that productivity gains are being systematically distorted in economies dependent on informational technologies. If that were so, productivity would look better in economies that were less reliant on these technologies. It doesn’t. Yet Talman and his co-authors are certainly right to argue that we need better methods of measurement for the 21st-century economy to capture both the upsides and downsides of new technologies.
The views expressed in this article are the author’s own and do not necessarily reflect Fair Observer’s editorial policy.