Moore’s Law and Sunk Costs: An E-government Dilemma

Moore’s Law,” first articulated in 1965, tells us that we will see a doubling in transistor capacity roughly every two years.  As predictions go, it has proven surprisingly durable, and is a handy conceptual framework for understanding why the internet is evolving so quickly.


Consider: when I was in college a decade ago, I carried around a spare floppy disk for saving files and transfering them from one computer to another.  It could store about 100 megabites, which was a lot more storage than you’d find on campus webmail.  A few friends of mine had new computers with a whole GIGAbite of space.  It wasn’t entirely clear to me what one would do with all that space.

Today, my iPod has 120 Gigs.  I save my files to gmail and they reside in the clouds.  Wireless networked computing means I can access them anywhere, even on the road from an iPhone.  My computer, bought 4 years ago, is antiquated and slow.  It’s a mac that doesn’t even have embedded video.  And therein lies a disconnect for businesses and governments in the digital age.

Let’s say that the University I work at decides to invest in a complete upgrade of their computer systems.  This includes hardward and software, a new fleet of computers; custom-built e-learning tools.  (full disclosure: this post was inspired by the “MyCourses” e-learning training I attended this morning)  In so doing, they incur some substantial sunk costs.  Those computers and those software programs have got to be useable for several years.  Organizations and governments don’t move seamlessly up the line presented in the figure above.  They move in a stepwise fashion, investing in new tools every few years which they are in turn saddled with as the technology continues to evolve.

Why is this important?  Well for one thing, it helps to explain why large bureacracies will virtually always have outdated websites.  Purchasing a fleet of new computers or moving into a new area of web-related activity caries a cost, and no organization can afford to keep pace with the rapid expansion.  Sunk costs are a reality of egovernment and organizational adoption.

Particularly in terms of software, however, it seems to me to be a strong argument in favor of partnering with organizations like Google or, better yet, relying on open source software platforms.  Learning today about Brown University’s “My Courses” software platform (it is similar to the “blackboard” site at Penn… a password protected site for students to access the syllabus, readings, grades, etc), I couldn’t help but think that the more advanced functionalities could be accomplished for free with a Ning site.  I’m sure “My Courses” was cutting edge when Brown first made the investment, [well… I’m not sure, but I’m willing to believe it] but companies like Ning have been free to innovate — incentivized to do so, in fact — while Brown has incurred the sunk costs.

In a digital environment that evolves as quickly as the internet, we can draw a hardline distinction between the entities who have an incentive to continually create and innovate (tech giants, startups who hope to challenge tech giants, and open source communities) and the entities who proceed in stepwise fashion, incurring sunk costs, staying pat for several years, and then incurring another round of sunk costs.  It then follows that one of the benefits afforded to informal networks (policy networks, activist networks, political networks, communities-of-interest… Clay Shirky’s “organizing without organizations”) is that their lack of resources also leads to a lack of sunk costs, meaning that any online action they engage in is likely to be, and essentially remain at, the cutting edge.

Those are just a few early thoughts on the topic.  E-gov hasn’t been one of my major research areas thus far, but I’m warming up to it.  Thoughts/reactions/feedback would be appreciated (as always)

One thought on “Moore’s Law and Sunk Costs: An E-government Dilemma

  1. Pingback: Book Blogging: Moore’s Law and Politics « shouting loudly

Comments are closed.