This uses ACS input tables from iPUMS... the variables that you need are:

• "State", "PUMA", "MIGPLAC1", "MIGPUMA" Or any variables that you want to filter and create demographic filters by.

There's a slight discrepancy from previous year and current year location, this is noted and mostly affects some counties in Alabama, IPUMS has chosen to bury its cross-walk table while it makes the correction, but you can still find it and its still good for most of the country.

The second cross-walk that's required is a PUMA geography to county one, the University of Missouri State Data Center web page has a set of wonderful tools to help with this translation depending on what type of geographic roll-up is required, areal weighting (apportioning the PUMA or division over multiple areas) but keep that in mind depending on the area that you want to roll up to, especially if you need to apportion a PUMA across METROS that cross state boundaries for example. Census and geographic data can get complicated!

Some quick visualizations - there are honestly so many: using the R circalize package

This chart filters out migration cells with less than 4000 households, just to simplify the data. There is lots of migration between Los Angeles and the Inland Empire, but not as much between the Inland-Empire and the Bay Area, San Diego is largely self-contained.

Other ways of visualizaing this data: sankey/ flow chart types which are better for one-way or net flows. In general migration data can get very busy, so it helps to consolidate areas into properly generalized groupings. For example, inner city to suburbs, to out-of-state flows depending on what you want to emphasize.

I also tried my hand at "great (well maybe not so great?) circle" gis maps using QGIS with Anita Grasser (sp?)'s flow package. It takes some work to make it look good, but it's much easier to use this to show net-migration change by arrow-thickness and shape.

The general pattern for migration in the bay area, is from the 6 or 7 central bay counties into peripheral counties as people get priced out of the bay. Then, as people get priced out of the outer-bay counties they simply move out of the state.

These charts still get busy, animated charts could help simplify some of that.

Gu, Z. (2014) circlize implements and enhances circular visualization in R. Bioinformatics. DOI: 10.1093/bioinformatics/btu393

IPUMS cite - Ruggle 2014