Am I doing something wrong with this?: My three attempts to download 66 ACS variables (for Stata) just for 10 largest US cities, selected using the special characteristics option, gives me just seven cities’ variables. Three – Dallas, San Jose, San Diego – are consistently missing, while the seven others arrive fine. Even an attempt to download the vars separately for Dallas, San Jose, and San Diego resulted in 0 variables downloaded. What am I doing wrong? (I am selecting 66 vars and 5 samples: 2015 1%, 2010-14 5%, 2005-09 5%, 2005 5%, 2000).
This is a frequent challenge. Basically, in source PUMS files since 1990, the most detailed geographic information available is the PUMA (Public Use Microdata Area), which are designed to contain at least 100,000 residents. PUMAs are sometimes coterminous with city boundaries, but they frequently encompass multiple cities and occasionally straddle city boundaries. Therefore, for most cities, and even for cities that are very large (as you’ve found), it is impossible to identify the exact set of corresponding PUMS records. You can see this if you look at the codes tab of the CITY variable, for each of the samples you’ve included in your data extract there are zero records in Dallas, San Jose, and San Diego. You can read more information about this, as well as some suggestions for alternative methods, on the comparability tab of the CITY variable description. IPUMS has an interactive map, that allow users to see how PUMAs are laid out. Additionally, based on the type of analysis you are pursuing, the NHGIS resource may be of benefit.
Thanks Jeff, will check out your ideas.