Housing Data

Source: OECD
Graphs: EE

100 = "long term average"

A few statistical comments on these "ratios" that may not be obvious to people not familiar with using normalized quantities in statistics.

The beauty of ratios is that they are "dimensionless". So as long as both quantities are denominated in the same unit, the unit of measurement simply doesn't matter. The answer is the same in dollars, yen, rupees, barrels of oil, and heads of cows, and for point-in-time comparisons, we don't have to worry about inflation, etc.

The problem with these ratios is that the are emphatically not "normal" (for starters, they're always positive) and they have a tendency to be centered around a number, typically 1. You can't just perform regressions on these because the distributions tend to be strange (log-normal if you're lucky but mostly power-law-ish.)

Also, the extreme ends of these ratios tends to be meaningless. You can only meaningfully compare stuff near the "central tendency" (Think about Steve Job's house to Steve Job's income to understand what the EE means by that.)

However, none of that applies for this particular data since we're just graphing the central tendency over time.