Data for migration comes from the American Community Survey; county-to-county flow datasets experience a longer lag time than other standard datasets available in FactFinder. 5-year rolling average data was used for migration for all geographies, as the Census Bureau does not release 1-year annual data. Data is not available at any geography below the county level; note that flows that are relatively small on the county level are often within the margin of error. The metropolitan area comparison was performed for the nine-county San Francisco Bay Area, in addition to the primary MSAs for the nine other major metropolitan areas, by aggregating county data based on current metropolitan area boundaries. Data prior to 2011 is not available on Vital Signs due to inconsistent Census formats and a lack of net migration statistics for prior years. Only counties with a non-negligible flow are shown in the data; all other pairs can be assumed to have zero migration.
Given that the vast majority of migration out of the region was to other counties in California, California counties were bundled into the following regions for simplicity:
Bay Area: Alameda, Contra Costa, Marin, Napa, San Francisco, San Mateo, Santa Clara, Solano, Sonoma
Central Coast: Monterey, San Benito, San Luis Obispo, Santa Barbara, Santa Cruz
Central Valley: Fresno, Kern, Kings, Madera, Merced, Tulare
Los Angeles + Inland Empire: Imperial, Los Angeles, Orange, Riverside, San Bernardino, Ventura
Sacramento: El Dorado, Placer, Sacramento, Sutter, Yolo, Yuba
San Diego: San Diego
San Joaquin Valley: San Joaquin, Stanislaus
Rural: all other counties (23)
One key limitation of the American Community Survey migration data is that it is not able to track emigration (movement of current U.S. residents to other countries). This is despite the fact that it is able to quantify immigration, generally by continent of origin. Thus the Vital Signs analysis focuses primarily on net domestic migration, while still specifically citing in-migration flows from countries abroad based on data availability.