Logan's Run and measurement dysfunction
Thoughts on measurement dysfunction in software organizations, by analogy with controlling health care costs and the novel Logan's Run. (Audio voiceover now included.)
This post was first published on the agile Teams WordPress blog on March 11, 2010. As of July 12, 2024 all agile Teams content has moved here to Agile Analytics and Beyond. Minor edits have been made to support audio voiceover for the podcast.
#/media/File:Logans_run_movie_poster.jpg){kind=link}
The other day I heard a story on the radio about the US initiative for controlling health care costs. Someone had studied various health conditions and tried to measure which ones, if controlled, would reduce our country's total health care costs.
The radio report said that controlling obesity would reduce the total costs, but getting more people to quit smoking would actually increase total costs. Instead of people dying young from heart disease or lung cancer, they would live so much longer that total costs of care over their now-longer lifetimes would rise instead of falling. So, the reporter implied, we shouldn't encourage people to quit smoking; just let them die young.
Wow. Even if the reporter was being sarcastic (if so, I missed that tone), what a thing to say.
I'm not exactly a fan of smoking, but as someone who has lost beloved family members too young to both heart disease and lung cancer, this was offensive. What, we should as a nation tackle the challenge of obesity, but it's OK to let people die young from lung cancer and other smoking-affected diseases because it'll be CHEAPER?!
Wasn't that pretty much the plot of the dystopian novel Logan's Run which I read in high school? Logan's world has too-scarce resources, so they just don't let anyone get old and expensive to take care of. Everyone gets 21 good years (30 years in the movie) and that's it. Lastday.
(I did some homework and it turns out that - as is usual with medical studies and stories - the data is mixed on whether health care costs more for people who quit than for people who continue to smoke.)
On reflection, I realized that Logan's Run (a book I hadn't thought about for years) is actually a great illustration of the problem of 'measurement dysfunction' in software development. You get what you measure. If all you measure and care about is absolute total health care costs, then Logan's Run-type solutions really are optimal, right? But the clear problem with that set of measurements is that they don't consider the value of human life, of wisdom, of parents and grandparents and elders, of having more good productive years of life and love instead of fewer.
If all you measure is total software development costs, then by all means, have the work done by whoever is cheapest. If all you care about is schedule, then go ahead and push that product out the door on time with a ton of latent (or known) defects. You will get what you measure.
You will also likely 'get' some things you don't measure and probably don't really want, like rising technical debt and customer support/maintenance costs. Like a cancer, they will gradually kill your product, metastasize throughout your organization, and cause great pain for you and your people. Your product and company will live an inexpensive (at least until the end) - but short - life in your virtual Logan's World.
As for me, I'd rather keep looking for Sanctuary, working towards a more sustainable world, and trying to find ways to help diagnose and cure lung cancer.