Don't hate the algorithm, hate the data

Embrace the change

Photo by Harpreet Singh Grewal on Unsplash

Hello,

Welcome to Known Unknowns, a newsletter influenced by algorithms (but not written by one!)—and loving it.

Don’t Blame the Algos

I've noticed that people really hate algorithms, or at least they think they do. People blame them for everything: the spread of disinformation, expensive airfares, trouble finding love, and now, rising rents.

There was a time in my life when I programmed complex algorithms—or so I thought at the time. Now, they seem pretty simple compared to today’s. But much of our fear of algorithms comes from a lack of understanding about what they are and what they do. Algorithms are just programs written to find the best result (for a given objective) based on the data you feed them. Sophisticated or not, what we often dislike about algorithms is the data itself—it reflects the world we live in. Algorithms simply clarify what’s already there.

I wrote for Bloomberg about Kamala Harris’s plan to crack down on landlords who use algorithms to set rent. She blames algorithms for rising rents and supports a bill that would ban landlords from using services that provide software to analyze data from multiple landlords to optimize lease terms. I’m not sure how different this is from landlords looking at comps online and setting rent. It’s just more efficient. One study shows that using this software makes rents more sensitive to market conditions (rising faster in tight markets, offering bigger discounts when markets soften), which is exactly what you’d expect.

It's true that rents increased as more landlords started using algorithms, but the housing shortage also worsened during that time. Again, algorithms reflect and amplify the world we live in.

Demonizing algorithms comes from fear of technology and ignorance. Sure, if someone has nefarious motives—like amplifying disinformation or colluding on rents—algorithms can be a tool to achieve those ends, but many tools could serve the same purpose. There's no evidence that landlords are colluding or even have the ability or incentive to do so. We don’t need to ban algorithms; they’re just another form of technology that often improves our lives, with the potential to do even more in the future. Instead, we need to address the root of the problem, a housing shortage or the objective of the algorithms’ owner.

Frankly, there are few good options these days when it comes to economic policy. It feels like every day Trump announces another plan to exempt some form of income or expense from taxes. There's no rhyme or reason to what he’s picking—and if it all comes to pass, we’ll need some very sophisticated algorithms just to do our taxes.

Unions Need to Modernize

Both Harris and Vance seem to want to follow Biden's lead and restore union power to its 1970s peak. The strikes we've seen over the past few weeks show how this idea is not only inconsistent with today’s economy but also with where it's headed.

For instance, one of the port workers’ demands (still being negotiated, even though they're back at work) was to limit or even ban automation. Boeing wants to bring back defined-benefit (DB) pensions, which in the past reduced mobility, created a financial drag, and limited resources for innovation. And unions involved in public works projects often drive up costs and extend timelines to absurd degrees. It seems like unions are largely resisting the modern economy, which isn't just bad for the economy and future prosperity—it also doesn't serve their members well.

The nature of work has changed, and so have the returns on being good at your job. The old union model, which compresses wages among workers in exchange for stability, doesn't suit the modern worker. This might explain why so many union drives fail, and why—despite polls showing broad support for unions—few people (aside from graduate students and journalists) actually want to join one.

Yet, workers still feel insecure, and many are shut out of the benefits of the growing economy. Unions have the potential to help them, but to remain relevant, they need to evolve—just like everything else. Ideally, unions would function more like guilds, offering health insurance, negotiating basic wages and working conditions, collaborating with employers on re-training and apprenticeships, and promoting flexibility. Instead, they often introduce rigidity and resist change.

Nevertheless, politicians seem determined to elevate the old union model in today’s modern economy. It makes no sense at all. Maybe in the future algorithms will replace union bosses.

In Other News

More evidence that neoliberalism worked.

Until next time, Pension Geeks!

Allison