What values do you value?
On value collapse and the balance between metrics and gestalt.
Does the world feel flatter than it used to? I think so. Complexity is harder to find these days, as mediocre simplicity reigns supreme. Who cares if a book is well-written if it goes viral enough to sell? Film reviews hardly matter in a world where viewership can be tracked to the millisecond and artistic merit becomes secondary to eyeball traffic. Try to watch The Minecraft Movie. It feels like a ninety-minute compilation of TikToks, full of bright colors and sharp cuts without a scrap of nuance. The bar wasn't high for that movie, but I grew up playing Minecraft and still found it unwatchable. Yes, I am bitter.
Quality is dead. Long live metrics.
C. Thi Nyugen, a philosophy professor at the University of Utah, calls this dynamic value collapse – "when rich, subtle values are replaced by quantified versions of those values". Here's an example of value collapse: a fitness tracker describes a walk as a number of steps and a heart rate. Data is captured, but is the walk? Instead of a stroll through the world full of sensations and wandering thought, the numbers describe bland, mechanical movement.
Flatness appears when decisions are made based on these collapsed values. Hammer, meet nail. From the perspective of a fitness tracker, an optimal walk goes up and down an appropriately-graded hill a dozen times at the adequate pace for maintaining a Zone 2 heartrate. When Netflix optimizes for views over enjoyment, the result is toothless and generic content for the greatest common denominator. This creates shows that don't speak to anyone in particular, but that everyone will sit through to kill time. This happens in algorithmic feeds too: maximizing engagement over any other parameter results in addictive rabbit holes of extremism.
Metrics themselves aren't the cause of anything. To attempt to understand the world, we must measure the world. Reduction is inevitable, but care is crucial.
For social impact, measuring the right way matters and pitfalls of value collapse abound. Consider an organization addressing hunger: they might measure Number of Meals Served across a dozen community kitchens. But if that becomes the primary metric for the organization, what happens next? When past data is used as the basis for new programs, important nuance can be lost. A focus on quantity might neglect quality, (amount of locally or organically grown food served) or ignore the rate of return customers. Optimizing for Number of Meals Served means vending machines of microwaveable dinners, a solution that doesn't address anything but the shortest-term problem.
Every problem is multidimensional, so addressing a limited set of dimensions means creating a limited solution.
Instead, that same organization could track the following on a monthly basis:
- Meals Served
- Pounds of Fresh Produce Donated
- Pounds of Fresh Produce Given Away
- Percentage of Customers Returning
- Average Number of Visits Per Customer
- Average Satisfaction with Meals
Just by adding a few more indicators to their program evaluation, this organization now has a far clearer picture of their impact, and a window into more areas where they can help their clients and address hunger in their communities.
Value collapse is inevitable in measurement, but when paired with an organizational awareness of nuance and regular reevaluation, it doesn't have to be a problem. Organizations who choose the right values to value will find themselves with more holistic and meaningful results down the line.