Gross domestic product (GDP) is the basic measure of a country’s overall economic output based on the market value of all the goods and services the country produces. Most measures of economic performance used by government officials to inform their policies and decisions are based on GDP figures. But, many concerns have been raised about the adequacy of GDP-based measurements given the major structural changes that economies around the world have been going through over the past few decades.
GDP is essentially a measure of production. While suitable when economies were dominated by the production of physical goods, GDP does not adequately capture the growing share of services and the production of increasingly complex solutions that characterize advanced economies. Nor does it reflect important economic activity beyond production, such as income, consumption and living standards.
A few years ago, a Commission on the Measurement of Economic Performance and Social Progress led by Nobel-prize winning economists Joseph Stiglitz and Amartya Sen was convened to look at the limits of GDP as an indicator of economic performance and social progress.
“What we measure affects what we do; and if our measurements are flawed, decisions may be distorted . . .,” it noted in its report issued in September of 2009. “So too, we often draw inferences about what are good policies by looking at what policies have promoted economic growth; but if our metrics of performance are flawed, so too may be the inferences that we draw.”
The Commission recommended complementing classical measures of GDP and economic production with additional measurements that captured people’s well being, as well as factoring in measurements of sustainability to help reflect the evolution of the economy into the future.
Over the past decade, a new set of concerns have arisen with the evolution of the Internet-based digital economy. How do you measure the value of the explosive amount of free information goods available over the Internet, including Wikipedia articles, Google maps, Linux open source software and You Tube videos? “We know less about the sources of value in the economy than we did 25 years ago,” wrote economists Erik Brynjolfsson and Adam Saunders in MIT’s Sloan Review:
“We see the influence of the information age everywhere, except in the GDP statistics. More people than ever are using Wikipedia, Facebook, Craigslist, Pandora, Hulu and Google. Thousands of new information goods and services are introduced each year. Yet, according to the official GDP statistics, the information sector (software, publishing, motion picture and sound recording, broadcasting, telecom, and information and data processing services) is about the same share of the economy as it was 25 years ago - about 4%. How is this possible? Don’t we have access to more information than ever before?”
In Why it Matters that the GDP Ignores Free Goods, a talk delivered at the 2012 Techonomy conference, Brynjolfsson said that despite being in the midst of a major technology revolution, official government statistics don’t include the value of digital goods and you could thus conclude that the information sector has not grown at all since the1960s. “Obviously, there’s some major measurement problems in the way we keep our statistics, and that’s a real problem because, as the saying goes, you can’t manage what you don’t measure.”
In the talk, he mentions a few of the problems in measuring the value of digital goods. The first is that the marginal cost of delivering them over the Internet is pretty close to zero. While in some cases their economic model is based on advertising, in most cases users contribute their time, and develop digital content for nothing. We are spending more and more of our time consuming and developing digital goods than ever before. “We’ve more than doubled the amount of time we spend on these goods in the past five years.”
The problem is that GDP measures the total amount spent on these goods and services. If the price is zero, then “zero times any quantity is still zero. So you could have an enormous of explosion of bits or articles or whatever else. If they’re priced at zero, the statisticians in Washington do the math and, lo and behold, it comes out as a big fat zero contribution for our GDP.” Traditional metrics have not been adequate for the information economy because so much of the digital economy has been free.
Brynjolfsson proposes looking at how users spend their limited time and attention online as a possible solution to the measurement problem. “You can pay for something with dollars, but you can pay for something with anything else that has value as well. And when you choose to spend time on consuming a good, you are paying for that. You are paying a piece of your life. You are paying attention, if you will. And that is something that is in finite quantity.”
In The Attention Economy: Measuring the Value of Free Digital Services on the Internet, a draft paper written by Brynjolfsson and MIT postdoctoral fellow JooHee Oh, they introduce a framework for quantifying the value of users’ attention to online applications with very low cash prices. After doing the math and plugging in numbers, the annual welfare gain from all these free digital goods over the Internet averaged over the past ten years is roughly $300 billion or $1400 per person.
They applied the same model to calculate the attention value of television, another free good. The value for television is higher, because people still spend significantly more hours watching television than on the Internet. However, while the value added of television has leveled off and is starting to fall a bit, the attention value added by the Internet is still growing and expected to surpass that of television within the next decade.
“[V]alue creation and value capture are not the same thing. Our economy tends to measure value capture. If we’re going to get 21st century economic policy right, or even just correctly model what’s working and why, we have to start moving to a model that measures value creation rather than value capture.”
These are not academic concerns. When developing IBM’s Internet strategy in the mid-1990s, a major part of our work involved figuring out a business model that made sense for our clients and was financially sound, justifying the major investments IBM was making in its company-wide Internet initiative. It wasn’t easy. The business models of the new Internet products and services would not have been justified if we just looked at their direct revenues, which were relatively small in the early years of the initiative. Most of the value was captured by existing hardware and software products, which saw incremental revenues since they were now being accessed by a growing number of users over the Internet.
For example, customer self-service applications, - e.g., tracking the status of a shipped package, making hotel reservations and shopping online, - were among the most popular in the early days of e-business. By letting users directly access transactions and information over the Internet with their PCs and browsers, IBM clients were able to provide much better customer service without having to add extra personnel, and IBM was able to sell more hardware and software to handle the increased volumes.
The story was similar a few years later when IBM embraced Linux across all its offerings. We formed a Linux Technology Center, an IBM team of developers who worked closely with the Linux open source development community around the world. The bulk of the value of Linux to IBM was created by this small team of Linux developers, but most of the value was subsequently captured by the product and service businesses which saw incremental revenue by supporting Linux.
We treated the Internet and Linux as platforms for innovation across the whole company. And, while much of the value was created by the small teams focused on developing the new initiatives, the bulk of the value was captured by the existing IBM business units. In the early years of the initiatives, they saw incremental revenues from supporting the new capabilities with their existing offerings. And, over time, they came up with all kinds of new products and services based on the growing marketplace presence of the Internet and Linux.
GDP measurements, while helping us better manage the 20th century industrial economy, do not adequately reflect some of the most critical areas we need to better understand and manage in our 21st century economy, including services and quality-of-life, digital goods over the Internet, or value creation and value capture. If our metrics of economic performance are flawed, our policies and decisions will be distorted. We need a revolution in measurements to go along with our digital technology and digital economy revolutions.