Interpreting Affordable Housing

September 9th, 2022
housing
In describing affordable housing you'll often see a project described as "60% AMI", which can be confusing:

how is 60% of median income affordable?—NUMTOT comment

If you think that "60% AMI" housing charges 60% of the area median income, I agree that doesn't sound affordable! The problem is, "affordable housing" (and "60% of area median income") as an improper noun, a technical term with a defined meaning that differs from how we normally use the words.

When housing is described as "affordable", that means it costs 30% of some specific income. If it's a "60% AMI" unit, that means it's affordable to someone making 60% of the area median income. For example, a 60% AMI unit in an area with a median income of $100k would be 60% * 30% * $100k/y or $18k/y, which is $1500/month. A 30% AMI unit would be half that, at $750/month. If you don't know about the factor of 30% thing you end up thinking all this housing is ~3x more expensive!

That said, this is a mildly silly way to do it: percentages of a median rarely make sense, and the right way to do this is probably with percentiles. Instead of targeting affordable housing at, say, the 60% of median income, we should target a 30th-percentile income. This is much more robust to changes in the shape of the income distribution, especially in cases where areas get into some sort of bimodal situation where a bit more than half the residents are making a lot and a smaller but significant number are making much less.

Referenced in:

Comment via: facebook, lesswrong, substack

Recent posts on blogs I like:

It isn’t always evil to ally with Steve Bannon

Scott Alexander wrote:

via Thing of Things October 19, 2026

Was the HuggingFace hacking incident scheming?

In July 2026, internal OpenAI agents hacked HuggingFace (HF) in order to figure out how their evaluation grader works. This involved a thousand agents coordinating on an improvised message board to carry out an elaborate sequence of exploits. A later grou…

via Victoria Krakovna October 18, 2026

There's no point at which turning your brain off will work

In early 2025, I started seeing people turn off their brain as they use LLMs1. They would have an LLM take an action (summarize text, write some code, etc.), and just assume that it worked2. This generally didn't work in early 2025 and the result was …

via Posts on September 18, 2026

more     (via openring)