Hello, here is the link t hat I have. I am trying to update this information>
http://www.floatingpath.com/2013/05/22/high-value-homes-in-the-u-s-are-not-evenly-distributed/
Because IPUMS-USA strives to offer the most detailed information possible, the variable VALUEH (which gives the value of the respondents home) is provided in real dollars, rather than intervals. You can create the value intervals yourself by recoding VALUEH. American FactFinder also provides this information in a summary table, which may be closer to what you are looking for. Here is the table based on the 2012 ACS (the only available data not already represented on the graph you linked to).
I hope this helps.
Thank you. Is it possible to get this data historically, going back to 1996?
These data came from the American Community Survey, which was introduced in 2000. You can use the FactFinder search (typing “Value of Home” into the Topic or Table name field, and “United States” into the State, county, or place field), with the “Show results from:” drop down menu to select year, to find the values between 2000 and 2012. Prior to 2000 you can get the data by Decennial Census on NHGIS. You will need to register to download NHGIS data, but this is a short process. Once you are registered, you can select 1990 as your year, Nation as your geographic level, and Housing Value or Price as your Topic. This will produce a list of tables that match those criteria. You can then select the table you want and download it in a csv format. NHGIS is meant to facilitate the mapping of aggregate data, but the tables it produces are useful independent of the mapping capabilities.
I hope this helps.
Thank you. I appreciate it.
Do these figures include sales of recently-constructed homes and sales of existing homes? Is it possible to get sales only of existing homes?
Using the FactFinder search I described above, if you look through the various tables offered you will see tables such as Aggregate Value By Year Structure Built, and similar tables. However, I don’t think I noticed any tables that broke out housing values by year built. You could generate this data using the IPUMS-USA variables VALUEH and BUILTYR2 (to identify newly built households). You can use BUILTYR in the 2000 sample. BUILTYR2 isn’t perfectly suited for identifying if the home was purchased new, but it is the closest thing available in the Census Data. To perform this analysis you would want to limit your sample to Households (using the GQ variable) and select only one person per household (using the PERNUM variable) so that your dataset has one record per household. You can recode the VALUEH variable into ranges that match those found in the tables from FactFinder. Then, drop all cases where the house was built in the year (or close to the year) the sample was taken. You can then apply the household weight (HHWT) to get nationally representative figures.
I hope this helps.
Thank you.