In the meantime, however, the US government has created a new web site, data.gov, to serve as a central source for finding government-published data made available freely to the public.
And as it turns out, quite a bit of data is made available between each census by other government agencies. The IRS for example knows who moves to a different address, a different county or a different state each year, and as it turns out they make this data available each year.
The latest data available is for people who changed addresses in 2007-2008, in other words, people who filed taxes or filed for an exemption with a different address for 2008 income than they used for their 2007 declaration.
Both in-flows and out-flows are available, so you can see for a given county where all the people who moved there during the year were previously resident, or where all the people who left a given county that year moved to.
My first instinct, given the collapse of the US economy in late 2008, was to see which counties were losing taxpayers like so much unemployed investment banker ballast. But as it turns out, the map was very boring. There weren't massive moves from Manhattan to Alabama as a result of the financial meltdown. The biggest loser was Chattahoochee County, Georgia (for more reason than one obviously). But nearly everyone that left the county moved to neighbouring Muscogee County. I can only assume that county line was redrawn, a trailer park was relocated, or the locals suddenly realised that their county's name meant "discourse with a prostitute" in the language of today's youth and decided to go walkabout.
I'll take door number two.
In-flows turned out to be far more interesting than out-flows, because they show the places that people are moving to certain cities/counties from.
I ended up creating a series of maps this time, showing the counties of origin of people who moved to some of the major US cities between 2007 and 2008, and as a comparison also created the same map for two smaller cities, Sioux City and Columbus.
Each map comes in two versions. The "linear" colour scale and the "quantile" colour scale. Both maps display data that shows, for a given county of destination, which counties the migrants originally lived in in 2007.
To read these maps, say for New York, looking at a county that has a colour corresponding to 0.5%, you should be read "0.5% of all migrants moving to New York in 2007-2008 are from this county".
"Linear" scale maps give a good idea of where the raw majority of migrants are from. The darkest counties are the ones where a lot of people came from.
"Quantile" scale maps give a better idea of the spread of counties of origin. On this map, you'll see all the counties that have beween 0.001% and 0.1% much more easily.
People moving from Hawaii, Alaska, Puerto Rico and the other US territories, as well as from foreign countries, are taken into account in the calculation of the percentage, but are not represented on the maps.
How "lower 48" of me. Sarah Palin would be livid.
So, with just the above ado, I give you... MAPS!!
(As always, click to enlarge)
If any conclusion can be drawn from these maps, I'd say that looking at the linear scale ones especially, it's obvious that people who move to big US cities come principally from other big cities, or from the immediate suburbs. And people who move to smaller US cities move from the rest of the state.
I'm brewing up an idea at the moment. A web site that can generate maps on the fly. But not like the crappy ones government web sites serve up. Interesting, interactive ones that will for example let you choose a county out of the whole list of what exists and map the in- and out-flows by county of origin/destination on demand.
Watch this space!
