In my daily travels of the internet, I came across another1 unusually long time-series of home prices and rents for a bunch of US cities.
The Historical Housing Prices (HHP) Project at the Philadelphia Fed is a pretty neat data set that attempts tracking the price of housing for sale and for rent over the 20th century, by using information gleaned from the real estate sections of historical newspapers.
Given that it’s pretty rare to find data of this nature, I thought I’d write up a quick post exploring a couple aspects of the data that are hard to visualize using the tools provided on the Philadelphia Fed website.
A little clean up
The data contains time-series for 30 US cities, but not all the cities have data going back to the Wild West era.
This is probably why the 1948 year is chosen by the Philadelphia Fed as the point at which they indexed prices, since it’s the first year where data is available for every city.
But, I’m interested in seeing a horse race between cities starting at the earliest dates possible in the data. To do that, I’m going to have to dump a few cities that don’t have data going back that far.
Looking at the distribution of starting years in the data, there’s about 15 cities that start in 1890, with the other half of cities having later start dates.
For the sake of this post, I’m just going to focus on cities that have data from 1900 and prior, which leaves me with most of the available data, but sadly, excludes a few cities that would be interesting to look at2.
I’ll also re-index the data to the year 1900, so we can look at how prices and rents evolved over 120 years of data.
Giddy Up!
First up, let’s take a look at how the house prince index evolved.
It’s interesting to see how wars (WW1 and WW2) seem to have played an important role in keeping prices from growing the way they did post-WW2.
The data only go to 2006, so, it’s also interesting to note that at that point in time, San Diego, Portland, and San Francisco had experienced the greatest growth in prices, relative to the year 1900.
While we’re at it, let’s have a quick look at rents using the same methodology.
Again, we can see some impacts of WW1 and WW2 in these series, and the usual suspects still top the list in terms of growth rates since 1900.
In Sum
It’s always interesting to see such long time-series on prices, but it’s important to keep in mind these are aggregate price metrics that have a lot of issues in terms of accuracy.
The rent series seem a bit questionable to me given what we know about rent prices these days, and I’m not sure I’d use these data for any serious analysis without developing a much deeper understanding of them.
But, all in all, it’s pretty cool that someone took the time to put these together and open these data up to the world.