On the Utility FallacyMay 6th, 2019 · 24 comments
A few years ago, I wrote an article for the Harvard Business Review’s website about the excesses of email culture. In an effort to destabilize the perceived necessity of our current moment of hyperactive communication, I explored a thought experiment in which email was banished altogether and replaced with pre-scheduled office hours.
“Office hours might not work for every organization,” I wrote, “although, as I’ve argued, they would probably apply in more settings than you might at first assume.”
Given the semi-satirical undertones of this exploration, I gave a nod toward Swift in the article’s title: “A Modest Proposal: Eliminate Email.”
I’m bringing this up now because a reader recently pointed me to a Reddit thread from last month that discusses this older piece. Overall, the thread is varied and fascinating. I want to highlight here, however, a few comments that I think are representative of a general line of resistance I often encounter — usually from fellow engineering types — when I write negatively about new technologies:
- “…people seem to severely underestimate how valuable it is to search past conversations, not to mention having a timestamp of when assignments and decisions got made”
- “I am totally with you here. Email is THE SUPERIOR tool for communication. People simply lack the discipline to manage their inbox.”
- “Having lower cost, lower friction communication is an absolute positive development.”
These points are an example of what I’ve come to call the utility fallacy, which is the tendency, when evaluating the impact of a technology, to confine your attention to comparing the technical features of the new technology to what it replaced.
From this perspective, email is self-evidently better than the memos, voicemails, and fax machines that it superseded.
But as I’ve learned in my years thinking and writing about such issues, when it comes to consumer-facing technologies, the more important story is almost always how they end up mutating our socio-cultural dynamics.
No one argues, for example, that it’s better to send an email than a fax. But the modern knowledge worker now sends 125 business emails a day, which works out to one every 3.85 minutes — vastly more back-and-forth communication than what was common in the pre-email era. One could certainly argue that this new behavior is not “better” in any useful sense.
You could retort that knowledge workers are, en mass, acting stupidly when using this new tool, and if they’d simply talk to an enlightened engineer about inbox management we’d all be fine. But I find this explanation both unlikely and condescending.
More plausible is the hypothesis that the introduction of low-friction communication disrupted the finely-tuned dynamical system mediating inter-personal office interactions, creating both unpredictable and unfortunate results (c.f., Leslie Perlow’s work on the cycle of responsiveness).
I’ve noticed a similar conflict in recent discussions about social media. When viewed through the lens of the utility fallacy, these tools are unambiguously good at allowing people to connect with each other and share information with minimum effort.
But this observation misses the fact that almost everything interesting about our current struggles with social media concerns the impact of these tools on our lives beyond the screen.
The point too often missed in a cooly instrumentalist understanding of technology is that we don’t use these tools in a vacuum; we instead participate in complicated social systems that can careen in unforeseen directions when powerful new technological forces are introduced. Features are important, but they’re not the whole story.