Metrics, Models and Methods – Wired for Innovation

But what is it?

Tech policy poses a number of different challenges. The perhaps most interesting one is that we do not agree on what the subject is and how to measure it. What is technology? How does it impact things like our economy? In Wired for Innovation Erik Brynjolfsson and Adam Saunders try to attack that issue in a number of different ways, essentially saying that our old measure of, for example, GDP, fail to capture the majority of the values created by technology brings to the table. The perhaps most powerful example in the book is E-bay. Brynjolfsson and Saunders note that research shows that Ebay creates 4 dollars of consumer surplus value (notoriously hard to measure) per transaction, totalling, back in 2003, 7 billion USD. The actual captured value in GDP for Ebay the same year? 1 billion USD. New services and products may well create something like 7 to 10 times the value that is captured in our old measures. This will necessarily lead us to make decisions that are at best suboptimal, and at worst really hurtful. When the likes of Nicholas Carr argue that technology does not matter, or, in fact, that it is crippling us, he does so from these old measures (and not only of economics, but also from an old conception of what culture and thinking is).

Getting the numbers right thus needs to be a priority. And getting them right means new models for value creation and capture.


In Wired for Innovation the authors note that there are now new methods being developed for capturing these new values. Measuring consumer surplus by looking at time is one. Currently, in our GDP-model, the value of Internet Access is equal to the fees that consumers pay for internet access, they note, and nothing more. That is roughly 100 dollars a year and consumer, which then becomes the value economists will say that the Internet has to that consumer, too. Added up that makes 0.2 percent of GDP. That is the value our models have of the Internet.

The average time people spend on the Internet of their leisure time is 10 percent. That means that we spend 10 percent of our (free) time on something that is worth 0.2 percent of GDP. Not necessarily insane, right, but still thought provoking. Measuring time gives another estimate of worth however, and with this method the average value of internet access to a consumer is 3000 USD a year. 30 times the GDP-estimate. Now, all of these methods are in their infancy, but they are usable, and should be – as the authors argue – developed in much greater detail.


It is well worth while to follow the work in MIT on this issue, here.