Status:Closed    Asked:May 02, 2017 - 03:58 PM

Converting Dollar Amounts to Constant 1999 Dollars with CPI99

I have downloaded the single year files for 2013, 2014 and 2015 and want to convert 2013 and 2014 to 2015. I understand that I need to first convert 2013 and 2014 to 1999 dollars and then convert the 1999 dollars to 2015. However, when I do a crosstab of the variables "year" by "CPI99", I see that the conversion factor in the data set for 2013 is .715 and for 2014 it is .704 which do not seem consistent with the conversion factors listed on

If I am understanding this page correctly, for data year 2013 (income year 2012), the conversion factor would be .726 (not .715 as it shows in the dataset); and for data year 2014 (income year 2013), the conversion factor would be .715. (not .704 as I mentioned previosly)

To double check, I downloaded the single year files for 2000, 2001 and 2002, expecting the CPI for 2000 to be 1 as indicated on the conversion factor list, but when I run a crosstab again of the variables "year" by "CPI99", I get .967 for year 2000. From the conversion chart, .967 which would be for income year 2000 but not data year 2000. It seems to me data year 2000 would have a conversion of 1 since it is for income year 1999. Is this correct?

I thought the year variable indicated the data year, not the income year. Please let me know if I've misunderstood the year variable or the conversion factors listed on the website.

Many thanks!

Do you have the same question? Follow this Question

Staff Answer


Jeff Bloem


It is sometimes a bit confusing as the IPUMS USA and IPUMS CPS projects often have very similar but slightly different variables. It looks like you are using IPUMS USA data but are comparing them with the IPUMS CPS CPI99 table. If you look the IPUMS USA CPI99 table, you should see that the data align. The reason for the difference in the CPI99 variable between the CPS and ACS samples is the ACS reflects survey year as the same as income year. In the CPS ASEC samples, on the other hand, the survey year is a year ahead of the income year.

I hope this helps.


May 03, 2017 - 08:57 AM

Report it


Thank you Jeff! This is very helpful and clarifying!


May 03, 2017 - 11:20 AM

Report it


alternative to a 3-year file
When will racesing be available in the 2009-2013 and 2010-2014 five year ACS?
Labor Force Participation by Age for the year 2014 looks very different
Comparing CPI99 value of 2015 in IPUMS CPS and IPUMS ACS datasets
Login   |   Register

Recently Active Members

View More »

Share |